[model-inventory] Model alias inventory update - 2026-07-19

[github-actions]

Summary

Daily model inventory check for 2026-07-19. All major model families are well-covered by existing alias patterns. Pricing in models.json aligns with the GitHub Docs billing table for all active Copilot models. No alias mapping changes are required; however, there are 14 Copilot-served legacy/utility model IDs not yet tracked in models.json.

• Providers queried: OpenAI, Anthropic, Gemini, Copilot (via reflect)
• Total models found: Copilot 36 · Anthropic 9 · Gemini 50 · OpenAI 219 = 314 (deduplicated cross-provider)
• Proposed alias changes: 0 — existing patterns provide full coverage
• Pricing gaps found: 14 Copilot-served models absent from models.json (legacy/utility category, low priority)

Provider Model Counts

Provider	Models Available	Status
openai	219	✅ ok
anthropic	9	✅ ok
gemini	50	✅ ok
copilot	36	✅ ok

Raw API Fields Discovered

Provider	Useful Fields
Anthropic	display_name, created_at, max_input_tokens, max_tokens, capabilities (batch, citations, code_execution, vision, streaming, tool_use)
Gemini	displayName, version, inputTokenLimit, outputTokenLimit, supportedGenerationMethods, thinking flag
OpenAI	id, object, created, owned_by — minimal metadata; no token limits or pricing
Copilot (reflect)	Model ID list only; no per-model billing object in reflect v1 endpoint

Notable: Anthropic's API now returns capabilities.vision, capabilities.tool_use, and capabilities.streaming flags suitable for enriching the vision and reasoning alias coverage checks in future runs. Gemini's thinking boolean and inputTokenLimit are reliable proxy signals for tier classification.

models.json Pricing Analysis

Pricing validation (docs table vs models.json)

All active Copilot-facing models in pkg/cli/data/models.json were validated against the GitHub Docs pricing page and found correct:

Model	Docs Input/M	Docs Output/M	models.json Input	models.json Output	Match
claude-haiku-4.5	$1.00	$5.00	1e-06	5e-06	✅
claude-sonnet-4	$3.00	$15.00	3e-06	1.5e-05	✅
claude-sonnet-4.5	$3.00	$15.00	3e-06	1.5e-05	✅
claude-sonnet-4.6	$3.00	$15.00	3e-06	1.5e-05	✅
claude-opus-4.5–4.8	$5.00	$25.00	5e-06	2.5e-05	✅
claude-sonnet-5	$2.00	$10.00	2e-06	1e-05	✅
claude-fable-5	$10.00	$50.00	1e-05	5e-05	✅
gemini-2.5-pro	$1.25	$10.00	1.25e-06	1e-05	✅
gemini-3-flash-preview	$0.50	$3.00	5e-07	3e-06	✅
gemini-3.5-flash	$1.50	$9.00	1.5e-06	9e-06	✅
gpt-5-mini	$0.25	$2.00	2.5e-07	2e-06	✅
gpt-5.3-codex	$1.75	$14.00	1.75e-06	1.4e-05	✅
gpt-5.4	$2.50	$15.00	2.5e-06	1.5e-05	✅
gpt-5.4-mini	$0.75	$4.50	7.5e-07	4.5e-06	✅
gpt-5.5	$5.00	$30.00	5e-06	3e-05	✅
gpt-5.6-luna	$1.00	$6.00	1e-06	6e-06	✅
gpt-5.6-sol	$5.00	$30.00	5e-06	3e-05	✅
gpt-5.6-terra	$2.50	$15.00	2.5e-06	1.5e-05	✅
mai-code-1-flash-picker	$0.75	$4.50	7.5e-07	4.5e-06	✅
kimi-k2.7-code	$0.95	$4.00	9.5e-07	4e-06	✅
raptor-mini	$0.25	$2.00	2.5e-07	2e-06	✅

Note: claude-sonnet-5 is at promotional pricing ($2.00/$10.00) through August 31, 2026.

Missing from models.json (live Copilot serve, not tracked)

These model IDs are served by Copilot's live endpoint but absent from pkg/cli/data/models.json. All are legacy, versioned, or utility models — not pricing-critical for workflow alias resolution:

Model ID	Category	Notes
gpt-3.5-turbo	Legacy	Old generation, low priority
gpt-3.5-turbo-0613	Legacy versioned	Old generation
gpt-4	Legacy	Old generation
gpt-4-0613	Legacy versioned	Old generation
gpt-4-o-preview	Legacy	Pre-release GPT-4o
gpt-4.1-2025-04-14	Versioned	Date-pinned GPT-4.1
gpt-4o-2024-05-13	Versioned (deprecated)	Superseded by newer
gpt-4o-2024-08-06	Versioned (deprecated)	Superseded by newer
gpt-4o-2024-11-20	Versioned (deprecated)	Superseded by newer
gpt-4o-mini-2024-07-18	Versioned (deprecated)	Superseded by newer
text-embedding-3-small	Embedding	Not an LLM
text-embedding-3-small-inference	Embedding	Not an LLM
text-embedding-ada-002	Embedding	Not an LLM
trajectory-compaction	Internal utility	Not user-facing

Recommendation: These can remain untracked. They are not referenced by any alias pattern and the deprecated GPT-4/4o-mini models are already in the exclusion list.

Historical entries in models.json not currently in live Copilot reflect

Model ID	Notes
claude-fable-5	In billing docs — may be available via non-standard routing
claude-sonnet-4	In billing docs — confirmed active
gemini-3.1-pro-preview	Not in reflect but listed in docs ($2.00/$12.00)
gemini-3.5-flash	Not in reflect but listed in docs ($1.50/$9.00)
gpt-5.2	Historical entry, not in reflect
gpt-5.2-codex	Historical entry, not in reflect
kimi-k2.7-code	In billing docs — may route through non-standard endpoint
raptor-mini	In billing docs — may route through non-standard endpoint

These should be retained as historical records.

Inferred vs stored pricing discrepancies

None found. All stored pricing entries match the GitHub Docs billing table exactly.

Alias Coverage Assessment

All 36 live Copilot model IDs and all live Anthropic/Gemini/OpenAI model IDs checked against pkg/workflow/data/model_aliases.json patterns. Zero uncovered models — every live model matches at least one alias (or any).

Key coverage confirmed:

• claude-sonnet-5 → sonnet ✅
• claude-opus-4.5–4.8 → opus ✅
• gpt-5.6-luna/sol/terra → gpt-5.6 ✅
• gemini-3.5-flash → gemini-flash, gemini-3.5-flash ✅
• mai-code-1-flash-picker → mai-code-1-flash-picker, mai-code ✅

Proposed Alias Updates

No alias changes are required for this cycle. All live model families are covered.

Full Model Lists by Provider

Copilot (36 models via reflect)

claude-haiku-4.5, claude-opus-4.5, claude-opus-4.6, claude-opus-4.7, claude-opus-4.8, claude-sonnet-4.5, claude-sonnet-4.6, claude-sonnet-5, gemini-2.5-pro, gemini-3-flash-preview, gpt-3.5-turbo, gpt-3.5-turbo-0613, gpt-4, gpt-4-0613, gpt-4-o-preview, gpt-4.1, gpt-4.1-2025-04-14, gpt-4o, gpt-4o-2024-05-13, gpt-4o-2024-08-06, gpt-4o-2024-11-20, gpt-4o-mini, gpt-4o-mini-2024-07-18, 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, mai-code-1-flash-picker, text-embedding-3-small, text-embedding-3-small-inference, text-embedding-ada-002, trajectory-compaction

Anthropic (9 models)

claude-haiku-4-5-20251001, claude-opus-4-1-20250805, claude-opus-4-5-20251101, claude-opus-4-6, claude-opus-4-7, claude-opus-4-8, claude-sonnet-4-5-20250929, claude-sonnet-4-6, claude-sonnet-5

Gemini (50 models — notable ones)

antigravity-preview-05-2026, gemini-2.0-flash, gemini-2.0-flash-lite, gemini-2.5-flash, gemini-2.5-flash-image, gemini-2.5-flash-lite, gemini-2.5-pro, gemini-3-flash-preview, gemini-3-pro-image, gemini-3-pro-preview, gemini-3.1-flash-lite, gemini-3.1-flash-tts-preview, gemini-3.1-pro-preview, gemini-3.5-flash, gemma-4-26b-a4b-it, gemma-4-31b-it, imagen-4.0-generate-001, lyria-3-pro-preview, nano-banana-pro-preview, veo-3.1-generate-preview (+30 more)

OpenAI (219 models — notable ones)

gpt-5, gpt-5-codex, gpt-5-mini, gpt-5-nano, gpt-5-pro, gpt-5.1, gpt-5.2, 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, o1, o1-mini, o3, o3-mini, o3-pro, o4-mini, gpt-image-1, gpt-image-2 (+197 more including alpha/internal variants)

Notes

• The Copilot reflect endpoint returns a flat model ID list only; it does not expose per-model billing.multiplier fields in this environment. Pricing was validated against the GitHub Docs billing table instead.
• claude-sonnet-5 is at promotional pricing through August 31, 2026 — the stored value should be re-validated after that date.
• gpt-5.4-nano remains in the deprecated exclusion list per policy and was not evaluated.
• The stale copilot-prefixed patterns (e.g. copilot/o1*, copilot/gemini-omni*) match no live reflect model IDs but remain valid forward-looking patterns for if/when Copilot adds those models.

Generated by 📦 Daily Model Inventory Checker · 96.6 AIC · ⌖ 9.56 AIC · ⊞ 8.7K · ◷

• expires on Jul 25, 2026, 4:24 PM UTC-08:00