Merge branch 'main' into perf/swarm-detached-log-tail

commit: 029fea4e86edda82dd016a8d9286b34f3d5c7d09
author: Patric Eckhart · 28d ago · 2026-09-24 11:29 UTC
committer: GitHub · 28d ago · 2026-09-24 11:29 UTC
merge of: 8db8b1e + 9a5c565 · diff against the first parent
signature: GPG UnverifiedB5690EEEBB952194 The key is unknown

10 files · +155 -22

M README.md +3 -1


@@ -749,10 +749,12 @@ Use `/login` and pick api key to paste an AI Studio key. zot probes `/v1beta
749	749
750	750		> Free-tier rate limits. AI Studio's free tier has tight per-minute and per-day caps that vary by model: `gemini-2.5-pro` is the strictest (a few requests per minute, ~50 per day), Flash and Flash-Lite are far more generous. If a Pro turn 429s with `"You exceeded your current quota"` while Flash on the same key still works, you've hit the Pro free-tier RPD. Either switch to Flash for agent loops, or [enable billing](https://aistudio.google.com/app/apikey) on your AI Studio project to flip the same key from free to pay-as-you-go pricing (`$1.25/M` input, `$10/M` output for Pro).
751	751
752		-	Reasoning levels (`--reasoning off\|minimum\|low\|medium\|high\|xhigh\|max`, also configurable through `/reasoning` or in `/settings` as reasoning level) map differently per generation. `max` is a distinct opt-in tier above `xhigh`. GPT-5.6 and adaptive-thinking Claude models receive native `max`; unsupported providers clamp it to their highest accepted effort. Budget-based providers retain their provider/model caps. Gemini 3.x uses the `thinkingLevel` enum (`MINIMAL`/`LOW`/`MEDIUM`/`HIGH`), with Gemini-3-Pro pinned to `LOW` minimum and `HIGH` for any medium-or-higher request. `off` sends no reasoning config except on always-on thinking models. Gemini 2.0 models have no thinking config.
	752	+	Reasoning levels (`--reasoning off\|minimum\|low\|medium\|high\|xhigh\|max`, also configurable through `/reasoning` or in `/settings` as reasoning level) map differently per generation. `max` is a distinct opt-in tier above `xhigh`. GPT-5.6, GPT-6, and adaptive-thinking Claude models receive native `max`; unsupported providers clamp it to their highest accepted effort. Budget-based providers retain their provider/model caps. Gemini 3.x uses the `thinkingLevel` enum (`MINIMAL`/`LOW`/`MEDIUM`/`HIGH`), with Gemini-3-Pro pinned to `LOW` minimum and `HIGH` for any medium-or-higher request. `off` sends no reasoning config except on always-on thinking models. Gemini 2.0 models have no thinking config.
753	753
754	754		Claude Opus 5.5 is available through Anthropic (`--provider anthropic --model claude-opus-5-5`) and GitHub Copilot (`--provider github-copilot --model claude-opus-5.5`, subject to account availability). Both entries support a 1,000,000-token context window and up to 128,000 output tokens. Adaptive thinking is always on, with `low`, `medium`, `high`, `xhigh`, and `max` effort levels. The default, `off`, and `minimum` requests use `low` instead. The reasoning selector omits unsupported levels. Catalog prices per million tokens are $4 input, $20 output, $0.20 cache reads, and $5 cache writes. Copilot billing remains governed by your plan.
755	755
	756	+	GPT-6 Sol and GPT-6 Luna are available through OpenAI API keys, OpenAI Codex subscriptions, and GitHub Copilot. Select `gpt-6-sol` or `gpt-6-luna` with `--provider openai`, `openai-responses`, `openai-codex`, or `github-copilot`. Both use the Responses API, support up to 128,000 output tokens, and expose `low` through `max` reasoning. OpenAI and Codex entries use a 272,000-token context window, while Copilot entries use 1,000,000 tokens. Access through subscription providers depends on account availability and organization policy.
	757	+
756	758		You can add additional Gemini model IDs to `models.json` under the `google` provider.
757	759
758	760		### Gemini Enterprise Agent Platform (formerly Google Vertex AI)


@@ -749,10 +749,12 @@ Use `/login` and pick api key to paste an AI Studio key. zot probes `/v1beta
749			749
750		> Free-tier rate limits. AI Studio's free tier has tight per-minute and per-day caps that vary by model: `gemini-2.5-pro` is the strictest (a few requests per minute, ~50 per day), Flash and Flash-Lite are far more generous. If a Pro turn 429s with `"You exceeded your current quota"` while Flash on the same key still works, you've hit the Pro free-tier RPD. Either switch to Flash for agent loops, or [enable billing](https://aistudio.google.com/app/apikey) on your AI Studio project to flip the same key from free to pay-as-you-go pricing (`$1.25/M` input, `$10/M` output for Pro).	750		> Free-tier rate limits. AI Studio's free tier has tight per-minute and per-day caps that vary by model: `gemini-2.5-pro` is the strictest (a few requests per minute, ~50 per day), Flash and Flash-Lite are far more generous. If a Pro turn 429s with `"You exceeded your current quota"` while Flash on the same key still works, you've hit the Pro free-tier RPD. Either switch to Flash for agent loops, or [enable billing](https://aistudio.google.com/app/apikey) on your AI Studio project to flip the same key from free to pay-as-you-go pricing (`$1.25/M` input, `$10/M` output for Pro).
751			751
752	-	Reasoning levels (`--reasoning off\|minimum\|low\|medium\|high\|xhigh\|max`, also configurable through `/reasoning` or in `/settings` as reasoning level) map differently per generation. `max` is a distinct opt-in tier above `xhigh`. GPT-5.6 and adaptive-thinking Claude models receive native `max`; unsupported providers clamp it to their highest accepted effort. Budget-based providers retain their provider/model caps. Gemini 3.x uses the `thinkingLevel` enum (`MINIMAL`/`LOW`/`MEDIUM`/`HIGH`), with Gemini-3-Pro pinned to `LOW` minimum and `HIGH` for any medium-or-higher request. `off` sends no reasoning config except on always-on thinking models. Gemini 2.0 models have no thinking config.	752	+	Reasoning levels (`--reasoning off\|minimum\|low\|medium\|high\|xhigh\|max`, also configurable through `/reasoning` or in `/settings` as reasoning level) map differently per generation. `max` is a distinct opt-in tier above `xhigh`. GPT-5.6, GPT-6, and adaptive-thinking Claude models receive native `max`; unsupported providers clamp it to their highest accepted effort. Budget-based providers retain their provider/model caps. Gemini 3.x uses the `thinkingLevel` enum (`MINIMAL`/`LOW`/`MEDIUM`/`HIGH`), with Gemini-3-Pro pinned to `LOW` minimum and `HIGH` for any medium-or-higher request. `off` sends no reasoning config except on always-on thinking models. Gemini 2.0 models have no thinking config.
753			753
754		Claude Opus 5.5 is available through Anthropic (`--provider anthropic --model claude-opus-5-5`) and GitHub Copilot (`--provider github-copilot --model claude-opus-5.5`, subject to account availability). Both entries support a 1,000,000-token context window and up to 128,000 output tokens. Adaptive thinking is always on, with `low`, `medium`, `high`, `xhigh`, and `max` effort levels. The default, `off`, and `minimum` requests use `low` instead. The reasoning selector omits unsupported levels. Catalog prices per million tokens are $4 input, $20 output, $0.20 cache reads, and $5 cache writes. Copilot billing remains governed by your plan.	754		Claude Opus 5.5 is available through Anthropic (`--provider anthropic --model claude-opus-5-5`) and GitHub Copilot (`--provider github-copilot --model claude-opus-5.5`, subject to account availability). Both entries support a 1,000,000-token context window and up to 128,000 output tokens. Adaptive thinking is always on, with `low`, `medium`, `high`, `xhigh`, and `max` effort levels. The default, `off`, and `minimum` requests use `low` instead. The reasoning selector omits unsupported levels. Catalog prices per million tokens are $4 input, $20 output, $0.20 cache reads, and $5 cache writes. Copilot billing remains governed by your plan.
755			755
			756	+	GPT-6 Sol and GPT-6 Luna are available through OpenAI API keys, OpenAI Codex subscriptions, and GitHub Copilot. Select `gpt-6-sol` or `gpt-6-luna` with `--provider openai`, `openai-responses`, `openai-codex`, or `github-copilot`. Both use the Responses API, support up to 128,000 output tokens, and expose `low` through `max` reasoning. OpenAI and Codex entries use a 272,000-token context window, while Copilot entries use 1,000,000 tokens. Access through subscription providers depends on account availability and organization policy.
			757	+
756		You can add additional Gemini model IDs to `models.json` under the `google` provider.	758		You can add additional Gemini model IDs to `models.json` under the `google` provider.
757			759
758		### Gemini Enterprise Agent Platform (formerly Google Vertex AI)	760		### Gemini Enterprise Agent Platform (formerly Google Vertex AI)

M docs/providers.md +16 -16


@@ -114,10 +114,10 @@ use Chat Completions without `reasoning_effort`; their reasoning is managed by t
114	114		endpoint. All three routes send user/agent initiator metadata and image-request
115	115		headers where appropriate.
116	116
117		-	As of September 8, 2026, the catalog includes 27 public chat models: Claude
118		-	Fable 5/5.1, Haiku 4.5, Opus 4.7/4.8/4.8 Fast/5, Sonnet 5, Gemini
	117	+	As of September 22, 2026, the catalog includes 30 public chat models: Claude
	118	+	Fable 5/5.1, Haiku 4.5, Opus 4.7/4.8/4.8 Fast/5/5.5, Sonnet 5, Gemini
119	119		3.5/3.6/3.7/3.8 Flash, GPT-5 mini, GPT-5.3-Codex, GPT-5.4/5.4 mini/5.5,
120		-	GPT-5.6 Luna/Sol/Terra, GPT-6 Astra, Grok 4.5/4.6, Kimi K2.7 Code/K3,
	120	+	GPT-5.6 Luna/Sol/Terra, GPT-6 Astra/Luna/Sol, Grok 4.5/4.6, Kimi K2.7 Code/K3,
121	121		and MAI-Code-1-Flash/1.1-Flash. MAI-Code-1-Flash uses the public picker ID
122	122		`mai-code-1-flash-picker` and is scheduled to retire on September 10, 2026.
123	123

@@ -139,24 +139,24 @@ September 1, 2026, except for individual annual Copilot Pro and Pro+ subscribers
139	139		Eligible subscribers can add it through `$ZOT_HOME/models.json`.
140	140		See [GitHub's model retirement history and plan exceptions](https://docs.github.com/en/copilot/reference/ai-models/supported-models#model-retirement-history).
141	141
142		-	## GPT-6 Astra
	142	+	## GPT-6 models
143	143
144		-	Select `gpt-6-astra` with provider `openai`, `openai-responses`,
145		-	`openai-codex`, or `github-copilot`, for example:
	144	+	Select `gpt-6-astra`, `gpt-6-sol`, or `gpt-6-luna` with provider `openai`,
	145	+	`openai-responses`, `openai-codex`, or `github-copilot`, for example:
146	146
147	147		```bash
148		-	zot --provider github-copilot --model gpt-6-astra
	148	+	zot --provider github-copilot --model gpt-6-sol
149	149		```
150	150
151		-	Astra uses the Responses API and supports `low`, `medium`, `high`, `xhigh`,
152		-	and `max` reasoning. OpenAI API-key and Codex subscription entries use the
153		-	272,000-token short-context tier by default; the Copilot entry uses a
154		-	1,050,000-token context window. All entries include the 128,000-token output
155		-	limit and standard token prices, including the higher rates for prompts above
156		-	272,000 tokens. Subscription access depends on your
157		-	plan, rollout, and organization policy; catalog availability does not grant access.
158		-	See the [OpenAI model specifications](https://developers.openai.com/api/docs/models/gpt-6-astra)
159		-	and [GitHub Copilot availability](https://github.blog/changelog/2026-09-04-gpt-6-astra-is-generally-available-in-github-copilot/).
	151	+	All three use the Responses API, accept image input, support `low`, `medium`,
	152	+	`high`, `xhigh`, and `max` reasoning, and allow up to 128,000 output tokens.
	153	+	OpenAI API-key and Codex subscription entries use a 272,000-token context
	154	+	window. Copilot uses 1,050,000 tokens for Astra and 1,000,000 tokens for Sol
	155	+	and Luna. Standard per-million-token prices are $10 input and $50 output for
	156	+	Astra, $2 input and $10 output for Sol, and $0.10 input and $0.50 output for
	157	+	Luna. Prompts above 272,000 tokens use each model's long-context price tier.
	158	+	Subscription access depends on your plan, rollout, and organization policy.
	159	+	Catalog availability does not grant access.
160	160
161	161		## API-key providers
162	162


@@ -114,10 +114,10 @@ use Chat Completions without `reasoning_effort`; their reasoning is managed by t
114		endpoint. All three routes send user/agent initiator metadata and image-request	114		endpoint. All three routes send user/agent initiator metadata and image-request
115		headers where appropriate.	115		headers where appropriate.
116			116
117	-	As of September 8, 2026, the catalog includes 27 public chat models: Claude	117	+	As of September 22, 2026, the catalog includes 30 public chat models: Claude
118	-	Fable 5/5.1, Haiku 4.5, Opus 4.7/4.8/4.8 Fast/5, Sonnet 5, Gemini	118	+	Fable 5/5.1, Haiku 4.5, Opus 4.7/4.8/4.8 Fast/5/5.5, Sonnet 5, Gemini
119		3.5/3.6/3.7/3.8 Flash, GPT-5 mini, GPT-5.3-Codex, GPT-5.4/5.4 mini/5.5,	119		3.5/3.6/3.7/3.8 Flash, GPT-5 mini, GPT-5.3-Codex, GPT-5.4/5.4 mini/5.5,
120	-	GPT-5.6 Luna/Sol/Terra, GPT-6 Astra, Grok 4.5/4.6, Kimi K2.7 Code/K3,	120	+	GPT-5.6 Luna/Sol/Terra, GPT-6 Astra/Luna/Sol, Grok 4.5/4.6, Kimi K2.7 Code/K3,
121		and MAI-Code-1-Flash/1.1-Flash. MAI-Code-1-Flash uses the public picker ID	121		and MAI-Code-1-Flash/1.1-Flash. MAI-Code-1-Flash uses the public picker ID
122		`mai-code-1-flash-picker` and is scheduled to retire on September 10, 2026.	122		`mai-code-1-flash-picker` and is scheduled to retire on September 10, 2026.
123			123

@@ -139,24 +139,24 @@ September 1, 2026, except for individual annual Copilot Pro and Pro+ subscribers
139		Eligible subscribers can add it through `$ZOT_HOME/models.json`.	139		Eligible subscribers can add it through `$ZOT_HOME/models.json`.
140		See [GitHub's model retirement history and plan exceptions](https://docs.github.com/en/copilot/reference/ai-models/supported-models#model-retirement-history).	140		See [GitHub's model retirement history and plan exceptions](https://docs.github.com/en/copilot/reference/ai-models/supported-models#model-retirement-history).
141			141
142	-	## GPT-6 Astra	142	+	## GPT-6 models
143			143
144	-	Select `gpt-6-astra` with provider `openai`, `openai-responses`,	144	+	Select `gpt-6-astra`, `gpt-6-sol`, or `gpt-6-luna` with provider `openai`,
145	-	`openai-codex`, or `github-copilot`, for example:	145	+	`openai-responses`, `openai-codex`, or `github-copilot`, for example:
146			146
147		```bash	147		```bash
148	-	zot --provider github-copilot --model gpt-6-astra	148	+	zot --provider github-copilot --model gpt-6-sol
149		```	149		```
150			150
151	-	Astra uses the Responses API and supports `low`, `medium`, `high`, `xhigh`,	151	+	All three use the Responses API, accept image input, support `low`, `medium`,
152	-	and `max` reasoning. OpenAI API-key and Codex subscription entries use the	152	+	`high`, `xhigh`, and `max` reasoning, and allow up to 128,000 output tokens.
153	-	272,000-token short-context tier by default; the Copilot entry uses a	153	+	OpenAI API-key and Codex subscription entries use a 272,000-token context
154	-	1,050,000-token context window. All entries include the 128,000-token output	154	+	window. Copilot uses 1,050,000 tokens for Astra and 1,000,000 tokens for Sol
155	-	limit and standard token prices, including the higher rates for prompts above	155	+	and Luna. Standard per-million-token prices are $10 input and $50 output for
156	-	272,000 tokens. Subscription access depends on your	156	+	Astra, $2 input and $10 output for Sol, and $0.10 input and $0.50 output for
157	-	plan, rollout, and organization policy; catalog availability does not grant access.	157	+	Luna. Prompts above 272,000 tokens use each model's long-context price tier.
158	-	See the [OpenAI model specifications](https://developers.openai.com/api/docs/models/gpt-6-astra)	158	+	Subscription access depends on your plan, rollout, and organization policy.
159	-	and [GitHub Copilot availability](https://github.blog/changelog/2026-09-04-gpt-6-astra-is-generally-available-in-github-copilot/).	159	+	Catalog availability does not grant access.
160			160
161		## API-key providers	161		## API-key providers
162			162

M packages/provider/catalog_builtin.go +3 -1


@@ -146,7 +146,7 @@ var builtinCatalog = []Model{
146	146		{Provider: "fireworks", ID: "accounts/fireworks/routers/glm-5p1-fast", DisplayName: "GLM 5.1 Fast", ContextWindow: 202800, MaxOutput: 131072, Reasoning: true, PriceInput: 2.8, PriceOutput: 8.8, PriceCacheRead: 0.52, BaseURL: "https://api.fireworks.ai/inference"},
147	147		{Provider: "fireworks", ID: "accounts/fireworks/routers/kimi-k2p6-turbo", DisplayName: "Kimi K2.6 Turbo", ContextWindow: 262000, MaxOutput: 262000, Reasoning: true, PriceInput: 2, PriceOutput: 8, PriceCacheRead: 0.3, BaseURL: "https://api.fireworks.ai/inference"},
148	148		// ----- github-copilot -----
149		-	// Public chat models checked against GitHub's supported-model list on 2026-09-08.
	149	+	// Public chat models checked against GitHub's supported-model list on 2026-09-22.
150	150		// New model limits and token prices: https://models.dev/api.json (github-copilot).
151	151		// Copilot's fast Opus model ID comes from the inference API's available-model list.
152	152		{Provider: "github-copilot", ID: "claude-fable-5", DisplayName: "Claude Fable 5", ContextWindow: 1000000, MaxOutput: 128000, Reasoning: true, AdaptiveThinking: true, PriceInput: 10, PriceOutput: 50, PriceCacheRead: 1, PriceCacheWrite: 12.5, BaseURL: "https://api.individual.githubcopilot.com"},

@@ -178,6 +178,8 @@ var builtinCatalog = []Model{
178	178		{Provider: "github-copilot", ID: "gpt-5.6-terra", DisplayName: "GPT-5.6 Terra", API: APIResponses, ContextWindow: 400000, MaxOutput: 128000, Reasoning: true, PriceInput: 0, PriceOutput: 0, PriceCacheRead: 0, BaseURL: "https://api.individual.githubcopilot.com"},
179	179		{Provider: "github-copilot", ID: "gpt-5.6-luna", DisplayName: "GPT-5.6 Luna", API: APIResponses, ContextWindow: 400000, MaxOutput: 128000, Reasoning: true, PriceInput: 0, PriceOutput: 0, PriceCacheRead: 0, BaseURL: "https://api.individual.githubcopilot.com"},
180	180		{Provider: "github-copilot", ID: "gpt-6-astra", DisplayName: "GPT-6 Astra", API: APIResponses, ContextWindow: 1050000, MaxOutput: 128000, Reasoning: true, PriceInput: 10, PriceOutput: 50, PriceCacheRead: 1, PriceCacheWrite: 12.5, PriceTierInputTokens: 272000, PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25, BaseURL: "https://api.individual.githubcopilot.com"},
	181	+	{Provider: "github-copilot", ID: "gpt-6-sol", DisplayName: "GPT-6 Sol", API: APIResponses, ContextWindow: 1000000, MaxOutput: 128000, Reasoning: true, PriceInput: 2, PriceOutput: 10, PriceCacheRead: 0.2, PriceCacheWrite: 2.5, PriceTierInputTokens: 272000, PriceInputAbove: 4, PriceOutputAbove: 15, PriceCacheReadAbove: 0.4, PriceCacheWriteAbove: 5, BaseURL: "https://api.individual.githubcopilot.com"},
	182	+	{Provider: "github-copilot", ID: "gpt-6-luna", DisplayName: "GPT-6 Luna", API: APIResponses, ContextWindow: 1000000, MaxOutput: 128000, Reasoning: true, PriceInput: 0.1, PriceOutput: 0.5, PriceCacheRead: 0.01, PriceCacheWrite: 0.125, PriceTierInputTokens: 272000, PriceInputAbove: 0.2, PriceOutputAbove: 0.75, PriceCacheReadAbove: 0.02, PriceCacheWriteAbove: 0.25, BaseURL: "https://api.individual.githubcopilot.com"},
181	183		// ----- google -----
182	184		{Provider: "google", ID: "gemini-3-flash-preview", DisplayName: "Gemini 3 Flash Preview", ContextWindow: 1048576, MaxOutput: 65536, Reasoning: true, PriceInput: 0.5, PriceOutput: 3, PriceCacheRead: 0.05, BaseURL: "https://generativelanguage.googleapis.com/v1beta"},
183	185		{Provider: "google", ID: "gemini-3-pro-preview", DisplayName: "Gemini 3 Pro Preview", ContextWindow: 1048576, MaxOutput: 65536, Reasoning: true, PriceInput: 2, PriceOutput: 12, PriceCacheRead: 0.2, BaseURL: "https://generativelanguage.googleapis.com/v1beta"},

 	{Provider: "fireworks", ID: "accounts/fireworks/routers/glm-5p1-fast", DisplayName: "GLM 5.1 Fast", ContextWindow: 202800, MaxOutput: 131072, Reasoning: true, PriceInput: 2.8, PriceOutput: 8.8, PriceCacheRead: 0.52, BaseURL: "https://api.fireworks.ai/inference"},
 	{Provider: "fireworks", ID: "accounts/fireworks/routers/kimi-k2p6-turbo", DisplayName: "Kimi K2.6 Turbo", ContextWindow: 262000, MaxOutput: 262000, Reasoning: true, PriceInput: 2, PriceOutput: 8, PriceCacheRead: 0.3, BaseURL: "https://api.fireworks.ai/inference"},
 	// ----- github-copilot -----
-	// Public chat models checked against GitHub's supported-model list on 2026-09-08.
+	// Public chat models checked against GitHub's supported-model list on 2026-09-22.
 	// New model limits and token prices: https://models.dev/api.json (github-copilot).
 	// Copilot's fast Opus model ID comes from the inference API's available-model list.
 	{Provider: "github-copilot", ID: "claude-fable-5", DisplayName: "Claude Fable 5", ContextWindow: 1000000, MaxOutput: 128000, Reasoning: true, AdaptiveThinking: true, PriceInput: 10, PriceOutput: 50, PriceCacheRead: 1, PriceCacheWrite: 12.5, BaseURL: "https://api.individual.githubcopilot.com"},
 	{Provider: "github-copilot", ID: "gpt-5.6-terra", DisplayName: "GPT-5.6 Terra", API: APIResponses, ContextWindow: 400000, MaxOutput: 128000, Reasoning: true, PriceInput: 0, PriceOutput: 0, PriceCacheRead: 0, BaseURL: "https://api.individual.githubcopilot.com"},
 	{Provider: "github-copilot", ID: "gpt-5.6-luna", DisplayName: "GPT-5.6 Luna", API: APIResponses, ContextWindow: 400000, MaxOutput: 128000, Reasoning: true, PriceInput: 0, PriceOutput: 0, PriceCacheRead: 0, BaseURL: "https://api.individual.githubcopilot.com"},
 	{Provider: "github-copilot", ID: "gpt-6-astra", DisplayName: "GPT-6 Astra", API: APIResponses, ContextWindow: 1050000, MaxOutput: 128000, Reasoning: true, PriceInput: 10, PriceOutput: 50, PriceCacheRead: 1, PriceCacheWrite: 12.5, PriceTierInputTokens: 272000, PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25, BaseURL: "https://api.individual.githubcopilot.com"},
+	{Provider: "github-copilot", ID: "gpt-6-sol", DisplayName: "GPT-6 Sol", API: APIResponses, ContextWindow: 1000000, MaxOutput: 128000, Reasoning: true, PriceInput: 2, PriceOutput: 10, PriceCacheRead: 0.2, PriceCacheWrite: 2.5, PriceTierInputTokens: 272000, PriceInputAbove: 4, PriceOutputAbove: 15, PriceCacheReadAbove: 0.4, PriceCacheWriteAbove: 5, BaseURL: "https://api.individual.githubcopilot.com"},
+	{Provider: "github-copilot", ID: "gpt-6-luna", DisplayName: "GPT-6 Luna", API: APIResponses, ContextWindow: 1000000, MaxOutput: 128000, Reasoning: true, PriceInput: 0.1, PriceOutput: 0.5, PriceCacheRead: 0.01, PriceCacheWrite: 0.125, PriceTierInputTokens: 272000, PriceInputAbove: 0.2, PriceOutputAbove: 0.75, PriceCacheReadAbove: 0.02, PriceCacheWriteAbove: 0.25, BaseURL: "https://api.individual.githubcopilot.com"},
 	// ----- google -----
 	{Provider: "google", ID: "gemini-3-flash-preview", DisplayName: "Gemini 3 Flash Preview", ContextWindow: 1048576, MaxOutput: 65536, Reasoning: true, PriceInput: 0.5, PriceOutput: 3, PriceCacheRead: 0.05, BaseURL: "https://generativelanguage.googleapis.com/v1beta"},
 	{Provider: "google", ID: "gemini-3-pro-preview", DisplayName: "Gemini 3 Pro Preview", ContextWindow: 1048576, MaxOutput: 65536, Reasoning: true, PriceInput: 2, PriceOutput: 12, PriceCacheRead: 0.2, BaseURL: "https://generativelanguage.googleapis.com/v1beta"},

M packages/provider/copilot_catalog_test.go +1 -1


@@ -8,7 +8,7 @@ import (
8	8		func TestCopilotCatalogCoversPublicChatModels(t *testing.T) {
9	9		groups := map[string][]string{
10	10		APIAnthropicMessages: {"claude-fable-5", "claude-fable-5.1", "claude-haiku-4.5", "claude-opus-4.7", "claude-opus-4.8", "claude-opus-4.8-fast", "claude-opus-5", "claude-opus-5.5", "claude-sonnet-5"},
11		-	APIResponses: {"gpt-5-mini", "gpt-5.3-codex", "gpt-5.4", "gpt-5.4-mini", "gpt-5.5", "gpt-5.6-luna", "gpt-5.6-sol", "gpt-5.6-terra", "gpt-6-astra", "grok-4.5", "grok-4.6", "mai-code-1-flash-picker", "mai-code-1.1-flash"},
	11	+	APIResponses: {"gpt-5-mini", "gpt-5.3-codex", "gpt-5.4", "gpt-5.4-mini", "gpt-5.5", "gpt-5.6-luna", "gpt-5.6-sol", "gpt-5.6-terra", "gpt-6-astra", "gpt-6-luna", "gpt-6-sol", "grok-4.5", "grok-4.6", "mai-code-1-flash-picker", "mai-code-1.1-flash"},
12	12		APICompletions: {"gemini-3.5-flash", "gemini-3.6-flash", "gemini-3.7-flash", "gemini-3.8-flash", "kimi-k2.7-code", "kimi-k3"},
13	13		}
14	14		for api, ids := range groups {


@@ -8,7 +8,7 @@ import (
8		func TestCopilotCatalogCoversPublicChatModels(t *testing.T) {	8		func TestCopilotCatalogCoversPublicChatModels(t *testing.T) {
9		groups := map[string][]string{	9		groups := map[string][]string{
10		APIAnthropicMessages: {"claude-fable-5", "claude-fable-5.1", "claude-haiku-4.5", "claude-opus-4.7", "claude-opus-4.8", "claude-opus-4.8-fast", "claude-opus-5", "claude-opus-5.5", "claude-sonnet-5"},	10		APIAnthropicMessages: {"claude-fable-5", "claude-fable-5.1", "claude-haiku-4.5", "claude-opus-4.7", "claude-opus-4.8", "claude-opus-4.8-fast", "claude-opus-5", "claude-opus-5.5", "claude-sonnet-5"},
11	-	APIResponses: {"gpt-5-mini", "gpt-5.3-codex", "gpt-5.4", "gpt-5.4-mini", "gpt-5.5", "gpt-5.6-luna", "gpt-5.6-sol", "gpt-5.6-terra", "gpt-6-astra", "grok-4.5", "grok-4.6", "mai-code-1-flash-picker", "mai-code-1.1-flash"},	11	+	APIResponses: {"gpt-5-mini", "gpt-5.3-codex", "gpt-5.4", "gpt-5.4-mini", "gpt-5.5", "gpt-5.6-luna", "gpt-5.6-sol", "gpt-5.6-terra", "gpt-6-astra", "gpt-6-luna", "gpt-6-sol", "grok-4.5", "grok-4.6", "mai-code-1-flash-picker", "mai-code-1.1-flash"},
12		APICompletions: {"gemini-3.5-flash", "gemini-3.6-flash", "gemini-3.7-flash", "gemini-3.8-flash", "kimi-k2.7-code", "kimi-k3"},	12		APICompletions: {"gemini-3.5-flash", "gemini-3.6-flash", "gemini-3.7-flash", "gemini-3.8-flash", "kimi-k2.7-code", "kimi-k3"},
13		}	13		}
14		for api, ids := range groups {	14		for api, ids := range groups {

M packages/provider/discover_copilot_test.go +1 -1


@@ -97,7 +97,7 @@ func TestModelAvailabilityFiltersWithoutDeletingCatalog(t *testing.T) {
97	97		t.Error("empty availability did not hide models")
98	98		}
99	99		SetModelAvailability("github-copilot", nil)
100		-	if len(ModelsForProvider("github-copilot")) != 28 {
	100	+	if len(ModelsForProvider("github-copilot")) != 30 {
101	101		t.Error("clearing restriction did not restore catalog")
102	102		}
103	103		}


@@ -97,7 +97,7 @@ func TestModelAvailabilityFiltersWithoutDeletingCatalog(t *testing.T) {
97		t.Error("empty availability did not hide models")	97		t.Error("empty availability did not hide models")
98		}	98		}
99		SetModelAvailability("github-copilot", nil)	99		SetModelAvailability("github-copilot", nil)
100	-	if len(ModelsForProvider("github-copilot")) != 28 {	100	+	if len(ModelsForProvider("github-copilot")) != 30 {
101		t.Error("clearing restriction did not restore catalog")	101		t.Error("clearing restriction did not restore catalog")
102		}	102		}
103		}	103		}

M packages/provider/extra_models.go +10 -0


@@ -53,6 +53,16 @@ var supplementCatalog = []Model{
53	53		PriceInput: 10, PriceOutput: 50, PriceCacheRead: 1, PriceCacheWrite: 12.5,
54	54		PriceTierInputTokens: 272000,
55	55		PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25},
	56	+	{Provider: "openai-responses", ID: "gpt-6-sol", DisplayName: "GPT-6 Sol (Responses)", API: APIResponses,
	57	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
	58	+	PriceInput: 2, PriceOutput: 10, PriceCacheRead: 0.2, PriceCacheWrite: 2.5,
	59	+	PriceTierInputTokens: 272000,
	60	+	PriceInputAbove: 4, PriceOutputAbove: 15, PriceCacheReadAbove: 0.4, PriceCacheWriteAbove: 5},
	61	+	{Provider: "openai-responses", ID: "gpt-6-luna", DisplayName: "GPT-6 Luna (Responses)", API: APIResponses,
	62	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
	63	+	PriceInput: 0.1, PriceOutput: 0.5, PriceCacheRead: 0.01, PriceCacheWrite: 0.125,
	64	+	PriceTierInputTokens: 272000,
	65	+	PriceInputAbove: 0.2, PriceOutputAbove: 0.75, PriceCacheReadAbove: 0.02, PriceCacheWriteAbove: 0.25},
56	66		{Provider: "openai-responses", ID: "gpt-5", DisplayName: "GPT-5 (Responses)",
57	67		ContextWindow: 400000, MaxOutput: 128000, Reasoning: true,
58	68		PriceInput: 1.25, PriceOutput: 10.00, PriceCacheRead: 0.125},


@@ -53,6 +53,16 @@ var supplementCatalog = []Model{
53	PriceInput: 10, PriceOutput: 50, PriceCacheRead: 1, PriceCacheWrite: 12.5,	53		PriceInput: 10, PriceOutput: 50, PriceCacheRead: 1, PriceCacheWrite: 12.5,
54	PriceTierInputTokens: 272000,	54		PriceTierInputTokens: 272000,
55	PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25},	55		PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25},
		56	+	{Provider: "openai-responses", ID: "gpt-6-sol", DisplayName: "GPT-6 Sol (Responses)", API: APIResponses,
		57	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
		58	+	PriceInput: 2, PriceOutput: 10, PriceCacheRead: 0.2, PriceCacheWrite: 2.5,
		59	+	PriceTierInputTokens: 272000,
		60	+	PriceInputAbove: 4, PriceOutputAbove: 15, PriceCacheReadAbove: 0.4, PriceCacheWriteAbove: 5},
		61	+	{Provider: "openai-responses", ID: "gpt-6-luna", DisplayName: "GPT-6 Luna (Responses)", API: APIResponses,
		62	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
		63	+	PriceInput: 0.1, PriceOutput: 0.5, PriceCacheRead: 0.01, PriceCacheWrite: 0.125,
		64	+	PriceTierInputTokens: 272000,
		65	+	PriceInputAbove: 0.2, PriceOutputAbove: 0.75, PriceCacheReadAbove: 0.02, PriceCacheWriteAbove: 0.25},
56	{Provider: "openai-responses", ID: "gpt-5", DisplayName: "GPT-5 (Responses)",	66		{Provider: "openai-responses", ID: "gpt-5", DisplayName: "GPT-5 (Responses)",
57	ContextWindow: 400000, MaxOutput: 128000, Reasoning: true,	67		ContextWindow: 400000, MaxOutput: 128000, Reasoning: true,
58	PriceInput: 1.25, PriceOutput: 10.00, PriceCacheRead: 0.125},	68		PriceInput: 1.25, PriceOutput: 10.00, PriceCacheRead: 0.125},

M packages/provider/gpt6_astra_test.go +87 -0


@@ -6,6 +6,93 @@ import (
6	6		"testing"
7	7		)
8	8
	9	+	func TestGPT6SolAndLunaCatalog(t *testing.T) {
	10	+	models := []struct {
	11	+	id string
	12	+	input float64
	13	+	output float64
	14	+	cacheRead float64
	15	+	cacheWrite float64
	16	+	inputAbove float64
	17	+	outAbove float64
	18	+	readAbove float64
	19	+	writeAbove float64
	20	+	}{
	21	+	{"gpt-6-sol", 2, 10, 0.2, 2.5, 4, 15, 0.4, 5},
	22	+	{"gpt-6-luna", 0.1, 0.5, 0.01, 0.125, 0.2, 0.75, 0.02, 0.25},
	23	+	}
	24	+	for _, model := range models {
	25	+	for _, name := range []string{"openai", "openai-responses", "openai-codex", "github-copilot"} {
	26	+	t.Run(name+"/"+model.id, func(t *testing.T) {
	27	+	m, err := FindModel(name, model.id)
	28	+	if err != nil {
	29	+	t.Fatal(err)
	30	+	}
	31	+	wantContext := 272000
	32	+	if name == "github-copilot" {
	33	+	wantContext = 1000000
	34	+	}
	35	+	if m.API != APIResponses \|\| m.ContextWindow != wantContext \|\| m.MaxOutput != 128000 \|\| !m.Reasoning \|\| m.Speculative {
	36	+	t.Fatalf("unexpected model capabilities: %+v", m)
	37	+	}
	38	+	if m.PriceInput != model.input \|\| m.PriceOutput != model.output \|\| m.PriceCacheRead != model.cacheRead \|\| m.PriceCacheWrite != model.cacheWrite {
	39	+	t.Fatalf("unexpected standard prices: %+v", m)
	40	+	}
	41	+	if m.PriceTierInputTokens != 272000 \|\| m.PriceInputAbove != model.inputAbove \|\| m.PriceOutputAbove != model.outAbove \|\| m.PriceCacheReadAbove != model.readAbove \|\| m.PriceCacheWriteAbove != model.writeAbove {
	42	+	t.Fatalf("unexpected long-context prices: %+v", m)
	43	+	}
	44	+	wantLevels := []string{"", "low", "medium", "high", "xhigh", "max"}
	45	+	if got := AvailableReasoningLevels(m); !slices.Equal(got, wantLevels) {
	46	+	t.Fatalf("reasoning levels = %q, want %q", got, wantLevels)
	47	+	}
	48	+	})
	49	+	}
	50	+	}
	51	+	}
	52	+
	53	+	func TestGPT6SolAndLunaResponsesReasoning(t *testing.T) {
	54	+	copilot := NewGithubCopilotClient("test-token").(*copilotClient).router
	55	+	clients := map[string]*codexClient{
	56	+	"openai": NewOpenAIResponsesNamed("test-token", "", "openai").(renamedClient).inner.(codexClient),
	57	+	"openai-responses": NewOpenAIResponsesNamed("test-token", "", "openai-responses").(renamedClient).inner.(codexClient),
	58	+	"openai-codex": NewOpenAICodex("test-token", "test-account", "").(*codexClient),
	59	+	"github-copilot": copilot.byAPI[APIResponses].(*codexClient),
	60	+	}
	61	+	for name, client := range clients {
	62	+	for _, model := range []string{"gpt-6-sol", "gpt-6-luna"} {
	63	+	t.Run(name+"/"+model, func(t *testing.T) {
	64	+	wire, err := client.buildRequest(Request{Model: model, Reasoning: "max"})
	65	+	if err != nil {
	66	+	t.Fatal(err)
	67	+	}
	68	+	if wire.Model != model \|\| wire.Reasoning == nil \|\| wire.Reasoning.Effort != "max" {
	69	+	t.Fatalf("unexpected request: %+v", wire)
	70	+	}
	71	+	})
	72	+	}
	73	+	}
	74	+	}
	75	+
	76	+	func TestGPT6SolAndLunaCopilotDispatch(t *testing.T) {
	77	+	router := NewGithubCopilotClient("test-token").(*copilotClient).router
	78	+	completions := &routeCaptureClient{name: "github-copilot"}
	79	+	responses := &routeCaptureClient{name: "github-copilot"}
	80	+	router.fallback = completions
	81	+	router.byAPI[APIResponses] = responses
	82	+
	83	+	for _, model := range []string{"gpt-6-sol", "gpt-6-luna"} {
	84	+	stream, err := router.Stream(context.Background(), Request{Model: model})
	85	+	if err != nil {
	86	+	t.Fatal(err)
	87	+	}
	88	+	for range stream {
	89	+	}
	90	+	}
	91	+	if !slices.Equal(responses.models, []string{"gpt-6-sol", "gpt-6-luna"}) \|\| len(completions.models) != 0 {
	92	+	t.Fatalf("Responses models = %v, Completions models = %v", responses.models, completions.models)
	93	+	}
	94	+	}
	95	+
9	96		func TestGPT6AstraCatalog(t *testing.T) {
10	97		for _, name := range []string{"openai", "openai-responses", "openai-codex", "github-copilot"} {
11	98		t.Run(name, func(t *testing.T) {


@@ -6,6 +6,93 @@ import (
6	"testing"	6		"testing"
7	)	7		)
8		8
		9	+	func TestGPT6SolAndLunaCatalog(t *testing.T) {
		10	+	models := []struct {
		11	+	id string
		12	+	input float64
		13	+	output float64
		14	+	cacheRead float64
		15	+	cacheWrite float64
		16	+	inputAbove float64
		17	+	outAbove float64
		18	+	readAbove float64
		19	+	writeAbove float64
		20	+	}{
		21	+	{"gpt-6-sol", 2, 10, 0.2, 2.5, 4, 15, 0.4, 5},
		22	+	{"gpt-6-luna", 0.1, 0.5, 0.01, 0.125, 0.2, 0.75, 0.02, 0.25},
		23	+	}
		24	+	for _, model := range models {
		25	+	for _, name := range []string{"openai", "openai-responses", "openai-codex", "github-copilot"} {
		26	+	t.Run(name+"/"+model.id, func(t *testing.T) {
		27	+	m, err := FindModel(name, model.id)
		28	+	if err != nil {
		29	+	t.Fatal(err)
		30	+	}
		31	+	wantContext := 272000
		32	+	if name == "github-copilot" {
		33	+	wantContext = 1000000
		34	+	}
		35	+	if m.API != APIResponses \|\| m.ContextWindow != wantContext \|\| m.MaxOutput != 128000 \|\| !m.Reasoning \|\| m.Speculative {
		36	+	t.Fatalf("unexpected model capabilities: %+v", m)
		37	+	}
		38	+	if m.PriceInput != model.input \|\| m.PriceOutput != model.output \|\| m.PriceCacheRead != model.cacheRead \|\| m.PriceCacheWrite != model.cacheWrite {
		39	+	t.Fatalf("unexpected standard prices: %+v", m)
		40	+	}
		41	+	if m.PriceTierInputTokens != 272000 \|\| m.PriceInputAbove != model.inputAbove \|\| m.PriceOutputAbove != model.outAbove \|\| m.PriceCacheReadAbove != model.readAbove \|\| m.PriceCacheWriteAbove != model.writeAbove {
		42	+	t.Fatalf("unexpected long-context prices: %+v", m)
		43	+	}
		44	+	wantLevels := []string{"", "low", "medium", "high", "xhigh", "max"}
		45	+	if got := AvailableReasoningLevels(m); !slices.Equal(got, wantLevels) {
		46	+	t.Fatalf("reasoning levels = %q, want %q", got, wantLevels)
		47	+	}
		48	+	})
		49	+	}
		50	+	}
		51	+	}
		52	+
		53	+	func TestGPT6SolAndLunaResponsesReasoning(t *testing.T) {
		54	+	copilot := NewGithubCopilotClient("test-token").(*copilotClient).router
		55	+	clients := map[string]*codexClient{
		56	+	"openai": NewOpenAIResponsesNamed("test-token", "", "openai").(renamedClient).inner.(codexClient),
		57	+	"openai-responses": NewOpenAIResponsesNamed("test-token", "", "openai-responses").(renamedClient).inner.(codexClient),
		58	+	"openai-codex": NewOpenAICodex("test-token", "test-account", "").(*codexClient),
		59	+	"github-copilot": copilot.byAPI[APIResponses].(*codexClient),
		60	+	}
		61	+	for name, client := range clients {
		62	+	for _, model := range []string{"gpt-6-sol", "gpt-6-luna"} {
		63	+	t.Run(name+"/"+model, func(t *testing.T) {
		64	+	wire, err := client.buildRequest(Request{Model: model, Reasoning: "max"})
		65	+	if err != nil {
		66	+	t.Fatal(err)
		67	+	}
		68	+	if wire.Model != model \|\| wire.Reasoning == nil \|\| wire.Reasoning.Effort != "max" {
		69	+	t.Fatalf("unexpected request: %+v", wire)
		70	+	}
		71	+	})
		72	+	}
		73	+	}
		74	+	}
		75	+
		76	+	func TestGPT6SolAndLunaCopilotDispatch(t *testing.T) {
		77	+	router := NewGithubCopilotClient("test-token").(*copilotClient).router
		78	+	completions := &routeCaptureClient{name: "github-copilot"}
		79	+	responses := &routeCaptureClient{name: "github-copilot"}
		80	+	router.fallback = completions
		81	+	router.byAPI[APIResponses] = responses
		82	+
		83	+	for _, model := range []string{"gpt-6-sol", "gpt-6-luna"} {
		84	+	stream, err := router.Stream(context.Background(), Request{Model: model})
		85	+	if err != nil {
		86	+	t.Fatal(err)
		87	+	}
		88	+	for range stream {
		89	+	}
		90	+	}
		91	+	if !slices.Equal(responses.models, []string{"gpt-6-sol", "gpt-6-luna"}) \|\| len(completions.models) != 0 {
		92	+	t.Fatalf("Responses models = %v, Completions models = %v", responses.models, completions.models)
		93	+	}
		94	+	}
		95	+
9	func TestGPT6AstraCatalog(t *testing.T) {	96		func TestGPT6AstraCatalog(t *testing.T) {
10	for _, name := range []string{"openai", "openai-responses", "openai-codex", "github-copilot"} {	97		for _, name := range []string{"openai", "openai-responses", "openai-codex", "github-copilot"} {
11	t.Run(name, func(t *testing.T) {	98		t.Run(name, func(t *testing.T) {

M packages/provider/models.go +28 -0


@@ -415,6 +415,20 @@ var Catalog = []Model{
415	415		PriceTierInputTokens: 272000,
416	416		PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25,
417	417		},
	418	+	{
	419	+	Provider: "openai", ID: "gpt-6-sol", DisplayName: "GPT-6 Sol", API: APIResponses,
	420	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
	421	+	PriceInput: 2, PriceOutput: 10, PriceCacheRead: 0.2, PriceCacheWrite: 2.5,
	422	+	PriceTierInputTokens: 272000,
	423	+	PriceInputAbove: 4, PriceOutputAbove: 15, PriceCacheReadAbove: 0.4, PriceCacheWriteAbove: 5,
	424	+	},
	425	+	{
	426	+	Provider: "openai", ID: "gpt-6-luna", DisplayName: "GPT-6 Luna", API: APIResponses,
	427	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
	428	+	PriceInput: 0.1, PriceOutput: 0.5, PriceCacheRead: 0.01, PriceCacheWrite: 0.125,
	429	+	PriceTierInputTokens: 272000,
	430	+	PriceInputAbove: 0.2, PriceOutputAbove: 0.75, PriceCacheReadAbove: 0.02, PriceCacheWriteAbove: 0.25,
	431	+	},
418	432
419	433		// ---- OpenAI Codex / ChatGPT subscription backend ----
420	434		// Same model ids as the OpenAI family, but routed through the

@@ -468,6 +482,20 @@ var Catalog = []Model{
468	482		PriceTierInputTokens: 272000,
469	483		PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25,
470	484		},
	485	+	{
	486	+	Provider: "openai-codex", ID: "gpt-6-sol", DisplayName: "GPT-6 Sol", API: APIResponses,
	487	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
	488	+	PriceInput: 2, PriceOutput: 10, PriceCacheRead: 0.2, PriceCacheWrite: 2.5,
	489	+	PriceTierInputTokens: 272000,
	490	+	PriceInputAbove: 4, PriceOutputAbove: 15, PriceCacheReadAbove: 0.4, PriceCacheWriteAbove: 5,
	491	+	},
	492	+	{
	493	+	Provider: "openai-codex", ID: "gpt-6-luna", DisplayName: "GPT-6 Luna", API: APIResponses,
	494	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
	495	+	PriceInput: 0.1, PriceOutput: 0.5, PriceCacheRead: 0.01, PriceCacheWrite: 0.125,
	496	+	PriceTierInputTokens: 272000,
	497	+	PriceInputAbove: 0.2, PriceOutputAbove: 0.75, PriceCacheReadAbove: 0.02, PriceCacheWriteAbove: 0.25,
	498	+	},
471	499		}
472	500
473	501		// DefaultModel is used when the user does not specify one.


@@ -415,6 +415,20 @@ var Catalog = []Model{
415	PriceTierInputTokens: 272000,	415		PriceTierInputTokens: 272000,
416	PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25,	416		PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25,
417	},	417		},
		418	+	{
		419	+	Provider: "openai", ID: "gpt-6-sol", DisplayName: "GPT-6 Sol", API: APIResponses,
		420	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
		421	+	PriceInput: 2, PriceOutput: 10, PriceCacheRead: 0.2, PriceCacheWrite: 2.5,
		422	+	PriceTierInputTokens: 272000,
		423	+	PriceInputAbove: 4, PriceOutputAbove: 15, PriceCacheReadAbove: 0.4, PriceCacheWriteAbove: 5,
		424	+	},
		425	+	{
		426	+	Provider: "openai", ID: "gpt-6-luna", DisplayName: "GPT-6 Luna", API: APIResponses,
		427	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
		428	+	PriceInput: 0.1, PriceOutput: 0.5, PriceCacheRead: 0.01, PriceCacheWrite: 0.125,
		429	+	PriceTierInputTokens: 272000,
		430	+	PriceInputAbove: 0.2, PriceOutputAbove: 0.75, PriceCacheReadAbove: 0.02, PriceCacheWriteAbove: 0.25,
		431	+	},
418		432
419	// ---- OpenAI Codex / ChatGPT subscription backend ----	433		// ---- OpenAI Codex / ChatGPT subscription backend ----
420	// Same model ids as the OpenAI family, but routed through the	434		// Same model ids as the OpenAI family, but routed through the

@@ -468,6 +482,20 @@ var Catalog = []Model{
468	PriceTierInputTokens: 272000,	482		PriceTierInputTokens: 272000,
469	PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25,	483		PriceInputAbove: 20, PriceOutputAbove: 75, PriceCacheReadAbove: 2, PriceCacheWriteAbove: 25,
470	},	484		},
		485	+	{
		486	+	Provider: "openai-codex", ID: "gpt-6-sol", DisplayName: "GPT-6 Sol", API: APIResponses,
		487	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
		488	+	PriceInput: 2, PriceOutput: 10, PriceCacheRead: 0.2, PriceCacheWrite: 2.5,
		489	+	PriceTierInputTokens: 272000,
		490	+	PriceInputAbove: 4, PriceOutputAbove: 15, PriceCacheReadAbove: 0.4, PriceCacheWriteAbove: 5,
		491	+	},
		492	+	{
		493	+	Provider: "openai-codex", ID: "gpt-6-luna", DisplayName: "GPT-6 Luna", API: APIResponses,
		494	+	ContextWindow: 272000, MaxOutput: 128000, Reasoning: true,
		495	+	PriceInput: 0.1, PriceOutput: 0.5, PriceCacheRead: 0.01, PriceCacheWrite: 0.125,
		496	+	PriceTierInputTokens: 272000,
		497	+	PriceInputAbove: 0.2, PriceOutputAbove: 0.75, PriceCacheReadAbove: 0.02, PriceCacheWriteAbove: 0.25,
		498	+	},
471	}	499		}
472		500
473	// DefaultModel is used when the user does not specify one.	501		// DefaultModel is used when the user does not specify one.

M packages/provider/reasoning.go +2 -2


@@ -259,7 +259,7 @@ func OpenAICompatAnthropicEffort(level string) string {
259	259		}
260	260
261	261		// OpenAICodexReasoningEffort maps zot levels onto the Responses API effort
262		-	// enum. GPT-5.6 and GPT-6 Astra support native max; other models clamp max to xhigh.
	262	+	// enum. GPT-5.6 and GPT-6 support native max; other models clamp max to xhigh.
263	263		func OpenAICodexReasoningEffort(level, model string) string {
264	264		switch NormalizeReasoning(level) {
265	265		case "minimum", "low":

@@ -282,5 +282,5 @@ func OpenAICodexReasoningEffort(level, model string) string {
282	282
283	283		func supportsResponsesMaxEffort(model string) bool {
284	284		id := strings.ToLower(model)
285		-	return strings.HasPrefix(id, "gpt-5.6-") \|\| id == "gpt-6-astra"
	285	+	return strings.HasPrefix(id, "gpt-5.6-") \|\| id == "gpt-6-astra" \|\| id == "gpt-6-sol" \|\| id == "gpt-6-luna"
286	286		}


@@ -259,7 +259,7 @@ func OpenAICompatAnthropicEffort(level string) string {
259		}	259		}
260			260
261		// OpenAICodexReasoningEffort maps zot levels onto the Responses API effort	261		// OpenAICodexReasoningEffort maps zot levels onto the Responses API effort
262	-	// enum. GPT-5.6 and GPT-6 Astra support native max; other models clamp max to xhigh.	262	+	// enum. GPT-5.6 and GPT-6 support native max; other models clamp max to xhigh.
263		func OpenAICodexReasoningEffort(level, model string) string {	263		func OpenAICodexReasoningEffort(level, model string) string {
264		switch NormalizeReasoning(level) {	264		switch NormalizeReasoning(level) {
265		case "minimum", "low":	265		case "minimum", "low":

@@ -282,5 +282,5 @@ func OpenAICodexReasoningEffort(level, model string) string {
282			282
283		func supportsResponsesMaxEffort(model string) bool {	283		func supportsResponsesMaxEffort(model string) bool {
284		id := strings.ToLower(model)	284		id := strings.ToLower(model)
285	-	return strings.HasPrefix(id, "gpt-5.6-") \|\| id == "gpt-6-astra"	285	+	return strings.HasPrefix(id, "gpt-5.6-") \|\| id == "gpt-6-astra" \|\| id == "gpt-6-sol" \|\| id == "gpt-6-luna"
286		}	286		}

M packages/provider/reasoning_test.go +4 -0


@@ -16,6 +16,8 @@ func TestAvailableReasoningLevels(t *testing.T) {
16	16		{name: "responses", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-5.5"}, want: []string{"", "low", "medium", "high", "xhigh"}},
17	17		{name: "responses native max", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-5.6-sol"}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},
18	18		{name: "astra native max", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-6-astra"}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},
	19	+	{name: "sol native max", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-6-sol"}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},
	20	+	{name: "luna native max", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-6-luna"}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},
19	21		{name: "adaptive", model: Model{Reasoning: true, AdaptiveThinking: true}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},
20	22		{name: "adaptive compatible", model: Model{Reasoning: true, AdaptiveThinkingCompat: true}, want: []string{"", "high"}},
21	23		{name: "gemini pro", model: Model{Provider: "google", ID: "gemini-3-pro-preview", Reasoning: true}, want: []string{"", "low", "high"}},

@@ -136,6 +138,8 @@ func TestReasoningEffortMappings(t *testing.T) {
136	138		{"max", "gpt-5.5", "high", "max", "xhigh", 32768, "max"},
137	139		{"max", "gpt-6-astra", "high", "max", "max", 32768, "max"},
138	140		{"max", "GPT-6-ASTRA", "high", "max", "max", 32768, "max"},
	141	+	{"max", "gpt-6-sol", "high", "max", "max", 32768, "max"},
	142	+	{"max", "gpt-6-luna", "high", "max", "max", 32768, "max"},
139	143		{"max", "gpt-6-unknown", "high", "max", "xhigh", 32768, "max"},
140	144		}
141	145		for _, tc := range cases {


@@ -16,6 +16,8 @@ func TestAvailableReasoningLevels(t *testing.T) {
16	{name: "responses", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-5.5"}, want: []string{"", "low", "medium", "high", "xhigh"}},	16		{name: "responses", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-5.5"}, want: []string{"", "low", "medium", "high", "xhigh"}},
17	{name: "responses native max", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-5.6-sol"}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},	17		{name: "responses native max", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-5.6-sol"}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},
18	{name: "astra native max", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-6-astra"}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},	18		{name: "astra native max", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-6-astra"}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},
		19	+	{name: "sol native max", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-6-sol"}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},
		20	+	{name: "luna native max", model: Model{Reasoning: true, API: APIResponses, ID: "gpt-6-luna"}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},
19	{name: "adaptive", model: Model{Reasoning: true, AdaptiveThinking: true}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},	21		{name: "adaptive", model: Model{Reasoning: true, AdaptiveThinking: true}, want: []string{"", "low", "medium", "high", "xhigh", "max"}},
20	{name: "adaptive compatible", model: Model{Reasoning: true, AdaptiveThinkingCompat: true}, want: []string{"", "high"}},	22		{name: "adaptive compatible", model: Model{Reasoning: true, AdaptiveThinkingCompat: true}, want: []string{"", "high"}},
21	{name: "gemini pro", model: Model{Provider: "google", ID: "gemini-3-pro-preview", Reasoning: true}, want: []string{"", "low", "high"}},	23		{name: "gemini pro", model: Model{Provider: "google", ID: "gemini-3-pro-preview", Reasoning: true}, want: []string{"", "low", "high"}},

@@ -136,6 +138,8 @@ func TestReasoningEffortMappings(t *testing.T) {
136	{"max", "gpt-5.5", "high", "max", "xhigh", 32768, "max"},	138		{"max", "gpt-5.5", "high", "max", "xhigh", 32768, "max"},
137	{"max", "gpt-6-astra", "high", "max", "max", 32768, "max"},	139		{"max", "gpt-6-astra", "high", "max", "max", 32768, "max"},
138	{"max", "GPT-6-ASTRA", "high", "max", "max", 32768, "max"},	140		{"max", "GPT-6-ASTRA", "high", "max", "max", 32768, "max"},
		141	+	{"max", "gpt-6-sol", "high", "max", "max", 32768, "max"},
		142	+	{"max", "gpt-6-luna", "high", "max", "max", 32768, "max"},
139	{"max", "gpt-6-unknown", "high", "max", "xhigh", 32768, "max"},	143		{"max", "gpt-6-unknown", "high", "max", "xhigh", 32768, "max"},
140	}	144		}
141	for _, tc := range cases {	145		for _, tc := range cases {