[model-inventory] Model alias inventory update - 2026-07-22

[github-actions]

Summary

13 Copilot models missing from models.json, 9 historical entries no longer returned by live API, alias coverage is solid for all current live models.

• Providers queried: OpenAI, Anthropic, Gemini, Copilot (via reflect)
• Total models found: OpenAI 219, Anthropic 9, Gemini 50, Copilot 36
• Proposed alias changes: 0 (all live models already covered)
• Pricing gaps found: 13 missing models, 9 historical entries

Provider Model Counts

Provider	Models Available	Status
openai	219	✅ ok (API key available)
anthropic	9	✅ ok (API key available)
gemini	50	✅ ok (API key available)
copilot	36	✅ ok (via reflect endpoint)

Note: OpenAI, Anthropic, and Gemini API keys were configured. Copilot models were fetched from the AWF api-proxy /reflect endpoint at runtime. Playwright pricing extraction was skipped (Chrome sandbox not available in this environment) — treat billing multiplier validation as unavailable for this run.

Raw API Fields Discovered

Anthropic (/v1/models)

• id, display_name, created_at, type
• max_input_tokens (e.g. 1,000,000 for claude-sonnet-5, claude-opus-4.7+)
• max_tokens (output limit; 128k for sonnet-5, 64k for 4.5 variants)
• capabilities: batch, citations, code_execution, effort (low/medium/high/xhigh/max), image_input, pdf_input, structured_outputs, thinking (with subtypes adaptive)
• Notable: all current models support thinking (adaptive), image_input, effort levels

OpenAI (/v1/models)

• id, object, created (Unix timestamp), owned_by
• Minimal metadata in the list endpoint — no token limits or capability flags exposed

Gemini (/v1beta/models)

• name (prefixed models/), displayName, description, version
• inputTokenLimit, outputTokenLimit — useful for tier inference
• supportedGenerationMethods (e.g. generateContent, countTokens, batchGenerateContent)
• temperature, topP, topK, maxTemperature
• thinking (boolean) — present on gemini-2.5-flash, gemini-2.5-pro, gemma-4 variants

Copilot (reflect endpoint)

• id, vendor, capabilities.limits.max_context_window_tokens, capabilities.limits.max_output_tokens
• capabilities.supports.vision, capabilities.type
• billing.multiplier — authoritative ET multiplier for Copilot models
• Note: the reflect endpoint returned a plain model ID list for this run (no billing/capabilities metadata attached). The enrichment loop found no unconfigured endpoints with models_url needing hydration.

models.json Pricing Analysis

Missing from models.json

The following models appear in the Copilot reflect inventory but have no entry under github-copilot in pkg/cli/data/models.json. These are mostly legacy/versioned IDs and embedding models that Copilot now exposes.

Model ID	Provider	Notes
gpt-3.5-turbo	copilot	legacy; now available via Copilot
gpt-3.5-turbo-0613	copilot	legacy dated variant
gpt-4-0613	copilot	legacy dated variant
gpt-4-o-preview	copilot	legacy preview
gpt-4.1-2025-04-14	copilot	dated variant of gpt-4.1
gpt-4o-2024-05-13	copilot	dated variant
gpt-4o-2024-08-06	copilot	dated variant
gpt-4o-2024-11-20	copilot	dated variant
gpt-4o-mini-2024-07-18	copilot	dated variant
text-embedding-3-small	copilot	embedding model
text-embedding-3-small-inference	copilot	embedding model variant
text-embedding-ada-002	copilot	legacy embedding model
trajectory-compaction	copilot	special-purpose internal model

OpenAI direct (not yet in models.json):

New models in OpenAI API not currently tracked: gpt-4.1-mini, gpt-4.1-nano, gpt-5.4-pro, gpt-5.5-pro, gpt-5-nano, o3-pro, o1-pro, gpt-image-2, gpt-image-1.5, gpt-image-1-mini.

Anthropic direct (new IDs):

claude-opus-4-1-20250805 — new dated variant; covered by opus alias glob but not explicitly in models.json.

Historical entries not currently returned by live APIs

These model IDs appear in pkg/cli/data/models.json under github-copilot but are absent from the Copilot reflect inventory. Retain as historical records.

Model ID	Notes
claude-fable-5	not yet live in Copilot
claude-sonnet-4	older generation, superseded
gemini-3.1-pro-preview	preview superseded
gemini-3.5-flash	not in Copilot (available in Gemini API)
gpt-5.2	superseded by 5.3+
gpt-5.2-codex	superseded
gpt-5.4-nano	intentionally deprecated Copilot-facing ID
kimi-k2.7-code	not returned
raptor-mini	not returned

Inferred vs stored pricing discrepancies

Billing multiplier data was not available in the reflect response for this run (billing fields not populated). No multiplier discrepancies can be computed. Docs pricing table extraction was skipped due to Chrome sandbox unavailability. Recommend re-running with working Playwright for a full multiplier audit.

No OpenAI or Anthropic pricing APIs were checked (they require separate endpoint queries). Gemini token limits were used as tier proxies:

Model ID	Token Limit (in/out)	Tier Inference
gemini-2.5-pro	1M / 65k	high-tier (large)
gemini-2.5-flash	1M / 65k + thinking	mid-tier with reasoning
gemini-3-flash-preview	(in API)	flash tier
gemini-3.5-flash, gemini-3.6-flash	(in API)	flash tier
gemini-3.1-pro-preview	(in API)	pro tier

Proposed Alias Updates

All 36 Copilot live models are covered by at least one existing alias (at minimum the any catchall). All Anthropic, OpenAI, and Gemini model families are covered by the existing globs. No alias changes are required.

Specific coverage confirmation:

• All claude-*sonnet* → sonnet, sonnet-6x, any
• All claude-*haiku* → haiku, any
• All claude-*opus* → opus, any
• All gpt-5* patterns → gpt-5, versioned gpt-5.x, any
• All gemini-*flash* → gemini-flash, vision, any
• All gemini-*pro* → gemini-pro, any
• mai-code-1-flash-picker → mai-code, mai-code-1-flash-picker, any
• o1, o3, o4-mini variants → reasoning, any

No alias JSON changes proposed for this run.

Potential future additions to consider

When these models become available via Copilot, consider dedicated aliases:

• gemini-3.6-flash — currently falls under gemini-3-flash; a gemini-3.6-flash alias may help for explicit version pinning
• gpt-5.5-pro, gpt-5.4-pro — already covered by gpt-5-pro, gpt-5.5, gpt-5.4
• o3-pro, o1-pro — covered by reasoning; no new alias needed

Full Model Lists by Provider

Copilot (36 models from 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 from API)

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 from API)

antigravity-preview-05-2026, aqa,
deep-research-max-preview-04-2026, deep-research-preview-04-2026, deep-research-pro-preview-12-2025,
gemini-2.0-flash, gemini-2.0-flash-001, gemini-2.0-flash-lite, gemini-2.0-flash-lite-001,
gemini-2.5-computer-use-preview-10-2025, gemini-2.5-flash, gemini-2.5-flash-image,
gemini-2.5-flash-lite, gemini-2.5-flash-preview-tts, gemini-2.5-pro, gemini-2.5-pro-preview-tts,
gemini-3-flash-preview, gemini-3-pro-image, gemini-3-pro-image-preview, gemini-3-pro-preview,
gemini-3.1-flash-image, gemini-3.1-flash-image-preview, gemini-3.1-flash-lite,
gemini-3.1-flash-lite-image, gemini-3.1-flash-lite-preview, gemini-3.1-flash-tts-preview,
gemini-3.1-pro-preview, gemini-3.1-pro-preview-customtools,
gemini-3.5-flash, gemini-3.5-flash-lite, gemini-3.6-flash,
gemini-embedding-001, gemini-embedding-2, gemini-embedding-2-preview,
gemini-flash-latest, gemini-flash-lite-latest, gemini-omni-flash-preview, gemini-pro-latest,
gemini-robotics-er-1.5-preview, gemini-robotics-er-1.6-preview,
gemma-4-26b-a4b-it, gemma-4-31b-it,
imagen-4.0-fast-generate-001, imagen-4.0-generate-001, imagen-4.0-ultra-generate-001,
lyria-3-clip-preview, lyria-3-pro-preview, nano-banana-pro-preview,
veo-3.1-fast-generate-preview, veo-3.1-generate-preview

OpenAI (219 models — notable subset)

gpt-4, gpt-4-0613, gpt-4-turbo, gpt-4.1, gpt-4.1-mini, gpt-4.1-nano,
gpt-4o, gpt-4o-mini,
gpt-5, gpt-5-mini, gpt-5-nano, gpt-5-pro, gpt-5-codex,
gpt-5.1, gpt-5.1-codex, gpt-5.2, gpt-5.2-codex, gpt-5.3-codex,
gpt-5.4, gpt-5.4-mini, gpt-5.4-pro, gpt-5.5, gpt-5.5-pro,
gpt-5.6-luna, gpt-5.6-sol, gpt-5.6-terra,
gpt-image-1, gpt-image-1-mini, gpt-image-1.5, gpt-image-2,
o1, o1-pro, o3, o3-mini, o3-pro, o4-mini, o4-mini-deep-research

Notes

• Playwright unavailable: Chrome sandbox (setuid) is not configured in this Docker environment, so the Copilot docs pricing table (Step 2.5) could not be extracted. Billing multiplier validation is skipped for this run.
• Reflect billing data: The Copilot reflect endpoint returned a flat model ID list for this run (no billing.multiplier fields populated). Primary pricing validation was therefore unavailable.
• Historical entries retained: claude-fable-5, claude-sonnet-4, gemini-3.1-pro-preview, gemini-3.5-flash, gpt-5.2, gpt-5.2-codex, gpt-5.4-nano, kimi-k2.7-code, raptor-mini are absent from live inventories but kept in models.json as historical records.
• model_aliases.json: No changes needed. All 36 Copilot live models are covered by existing alias patterns.

Generated by 📦 Daily Model Inventory Checker · sonnet46 87.1 AIC · ⌖ 8.05 AIC · ⊞ 8.7K · ◷

• expires on Jul 28, 2026, 4:19 PM UTC-08:00

[github-actions]

This issue is being closed as outdated. A newer issue has been created: #47434

View newer issue

---

This action was performed automatically by the Daily Model Inventory Checker workflow.