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

[github-actions]

Summary

Daily model inventory check for 2026-07-07. Four providers were queried. All live models are
covered by existing alias patterns with two exceptions: gpt-5.6 is present in the OpenAI live
API but missing from models.json, and raptor-mini is in models.json (github-copilot
provider) but has no alias entry. The Copilot docs billing table was empty (pre-job scrape
returned no rows), so billing multiplier validation relies on the reflect endpoint, which
returns null billing for all models in this proxy configuration.

• Providers queried: OpenAI, Anthropic, Gemini, Copilot (reflect)
• Total models found: 310 (OpenAI: 219, Gemini: 50, Copilot reflect: 32, Anthropic: 9)
• Proposed alias changes: 1 (add raptor-mini alias)
• Pricing gaps found: 1 (gpt-5.6 missing from models.json openai provider)

Provider Model Counts

Provider	Models Available	Status
openai	219	✅ ok
anthropic	9	✅ ok
gemini	50	✅ ok
copilot (reflect)	32	✅ ok

Raw API Fields Discovered

OpenAI

Fields available per model: id, object, created, owned_by. No pricing or capability
metadata is returned from the list endpoint — only model IDs and ownership.

Anthropic

Fields available: id (dash-formatted), display_name. Family grouping (haiku/sonnet/opus)
inferred from name. No token limits or pricing fields in the list endpoint.

Gemini

Rich metadata per model: name, displayName, description, inputTokenLimit,
outputTokenLimit, supportedGenerationMethods, temperature, topP, topK,
maxTemperature, thinking (boolean). Context limits are useful as a pricing-tier signal.
Notable: all Gemini 3.x models report 1M input / 64K output limits regardless of Flash vs Pro
classification; the image/TTS variants have smaller windows (64K–128K input).

Copilot reflect (/reflect)

Fields in the endpoint object: provider, port, base_url, configured, models (list
of IDs), models_url. The billing field is null for all models — this proxy configuration
does not expose multiplier data. No vendor metadata is returned through the reflect endpoint's
models list; vendor data is only available by querying individual model details from the API.

models.dev

Per-model fields: id, name, description, family, attachment, reasoning,
reasoning_options, tool_call, structured_output, temperature, knowledge,
release_date, last_updated, modalities.input/output, open_weights, limit.context/input/output,
cost.input/output/cache_read (in $/M tokens). This is the richest source for baseline
pricing and capability metadata.

models.json Pricing Analysis

Missing from models.json

Model ID	Provider	Inferred Pricing	Basis
gpt-5.6	openai	~$30/$180 per M (similar to gpt-5.5-pro tier)	models.dev not indexed yet; family pattern suggests pro-tier
gpt-5.6-luna	openai	—	Sub-variant of gpt-5.6; defer to parent
gpt-5.6-sol	openai	—	Sub-variant of gpt-5.6; defer to parent
gpt-5.6-terra	openai	—	Sub-variant of gpt-5.6; defer to parent

Note: The alias gpt-5.6 already exists in model_aliases.json with patterns
copilot/gpt-5.6* and openai/gpt-5.6*. The gap is only in models.json pricing entries.
The model is not yet available through the Copilot reflect endpoint.

Historical entries not currently returned

The following model IDs appear in models.json but are absent from all live inventories.
These should remain as historical records:

• claude-3-5-haiku-20241022, claude-3-5-haiku-latest (Anthropic — older 3.5 generation)
• claude-3-5-sonnet-20240620, claude-3-5-sonnet-20241022 (Anthropic — older 3.5 generation)
• claude-3-7-sonnet-20250219 (Anthropic — 3.7 generation)
• claude-3-haiku-20240307, claude-3-sonnet-20240229, claude-3-opus-20240229 (Anthropic — Claude 3)
• claude-opus-4-0, claude-opus-4-1, claude-opus-4-20250514, claude-sonnet-4-0, claude-sonnet-4-20250514 (Anthropic — early 4.x)
• claude-fable-5 (Anthropic — only in models.json, not in live Anthropic API)
• gpt-5.1, gpt-5.1-chat-latest, gpt-5.1-codex, gpt-5.1-codex-max, gpt-5.1-codex-mini (OpenAI — older 5.x)
• gpt-5.2, gpt-5.2-codex (OpenAI — older 5.x)
• gpt-5.4-nano (OpenAI — intentionally deprecated Copilot-facing ID; excluded per task rules)
• gemini-3.1-pro-preview, gemini-3.5-flash (github-copilot — not in reflect; may be pre-deployment)
• raptor-mini (github-copilot — in models.json with pricing, but NOT in Copilot reflect; may be unreleased)
• kimi-k2.7-code (github-copilot — in models.json, NOT in Copilot reflect)

Inferred vs stored pricing discrepancies

No discrepancies found for models present in both live inventory and models.json. Copilot
reflect returns null billing for all models, so direct multiplier comparison is not possible.
The docs billing table (/copilot/reference/copilot-billing/models-and-pricing) returned an
empty models array this run.

Proposed Alias Updates

1. Add raptor-mini alias

What: New alias raptor-mini for the raptor-mini model in the github-copilot provider.

Why: raptor-mini is present in pkg/cli/data/models.json under github-copilot with
pricing data (input: $0.25/M, output: $2/M) and wire_api: completions, making it
addressable at runtime. However, it has no corresponding entry in model_aliases.json, so
workflow authors cannot reference it by a stable name. Its extremely low pricing puts it in
the same class as mini/small models.

Syntax:

{
  "raptor-mini": [
    "copilot/raptor-mini*",
    "openai/raptor-mini*"
  ]
}

Consider also adding raptor-mini to the mini alias chain as it falls in the lightweight tier.

2. Add gpt-5.6 to models.json

What: Add gpt-5.6 (and potentially gpt-5.6-luna, gpt-5.6-sol, gpt-5.6-terra) to
the openai provider in pkg/cli/data/models.json.

Why: gpt-5.6 is a stable model ID returned by the OpenAI live API. The gpt-5.6 alias
already exists in model_aliases.json, so runtime routing is configured, but models.json
pricing data is missing. Running make refresh-models-json should pull it from models.dev
when available.

Action: Run make refresh-models-json and verify the entry is added. If models.dev does
not yet index it, add a placeholder entry manually with cost inferred from gpt-5.5-pro
(same tier pattern: reasoning + pro-level).

Full Model Lists by Provider

Anthropic (9 live 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

OpenAI (stable non-dated models, 219 total)

chatgpt-image-latest
gpt-5, gpt-5.1, gpt-5.2, gpt-5.2-pro
gpt-5.3-codex, gpt-5.4, gpt-5.4-mini, gpt-5.4-nano, gpt-5.4-pro
gpt-5.5, gpt-5.5-pro
gpt-5.6, gpt-5.6-luna, gpt-5.6-sol, gpt-5.6-terra  ← NEW
o1, o1-pro, o3, o3-mini, o3-pro, o4-mini
o3-deep-research, o4-mini-deep-research
gpt-image-1, gpt-image-1-mini, gpt-image-1.5, gpt-image-2
text-embedding-3-large, text-embedding-3-small, text-embedding-ada-002
(+ many dated/experimental variants)

Copilot reflect (32 models)

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
mai-code-1-flash-picker
text-embedding-3-small, text-embedding-3-small-inference, text-embedding-ada-002

Gemini (50 models, key subset)

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-image-preview, gemini-3-pro-preview
gemini-3.1-flash-image, gemini-3.1-flash-lite, gemini-3.1-flash-lite-image, gemini-3.1-pro-preview
gemini-3.5-flash
gemini-omni-flash-preview
antigravity-preview-05-2026
deep-research-preview-04-2026, deep-research-max-preview-04-2026, deep-research-pro-preview-12-2025
nano-banana-pro-preview
veo-3.1-generate-preview, veo-3.1-fast-generate-preview, veo-3.1-lite-generate-preview
lyria-3-pro-preview, lyria-3-clip-preview
imagen-4.0-generate-001, imagen-4.0-ultra-generate-001, imagen-4.0-fast-generate-001
gemma-4-26b-a4b-it, gemma-4-31b-it
gemini-robotics-er-1.5-preview, gemini-robotics-er-1.6-preview
gemini-2.5-computer-use-preview-10-2025
(+ embedding, TTS, and other modality models)

Notes

• Alias coverage is comprehensive: All 32 Copilot reflect models are covered by existing
  alias patterns (*sonnet*, *opus*, *haiku*, gpt-5*, gemini-*flash*, gemini-*pro*,
  mai-code*). No new alias patterns are required for currently-deployed Copilot models.
• Gemini 3.x image models (gemini-3-pro-image, gemini-3.1-flash-image, etc.) are
  covered by the image-generation alias via google/gemini-*image*.
• gpt-5.6 alias patterns already exist in model_aliases.json but the model is not yet
  in Copilot reflect and has no models.json entry.
• Copilot billing table was empty this run — pre-job scrape returned "models": [].
  Billing multiplier validation was not possible.
• Deprecated IDs (gpt-4o-mini, gpt-4o, gpt-4.1, gpt-5.4-nano) intentionally
  excluded from gap analysis per task policy.

Generated by 📦 Daily Model Inventory Checker · 103.6 AIC · ⌖ 18.4 AIC · ⊞ 7.8K · ◷

• expires on Jul 13, 2026, 4:27 PM UTC-08:00

[github-actions]

This issue is being closed as outdated. A newer issue has been created: #44139

View newer issue

---

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