[model-inventory] Model alias inventory update - 2026-06-22

[github-actions]

Summary

Daily model alias inventory run for 2026-06-22. One new alias is proposed (search). Several new model families (Gemini 3.1-x, Gemma 4, Imagen 4) were confirmed fully covered by existing wildcard patterns. No pricing discrepancies found for tracked models. make refresh-models-json failed (HTTP 403 from models.dev) — pre-fetched models-dev/api.json used for cross-validation instead.

• Providers queried: OpenAI, Anthropic, Gemini, Copilot
• Total models found: 296 (OpenAI: 208, Anthropic: 8, Gemini: 50, Copilot: 30)
• Proposed alias changes: 1 (search)
• Pricing gaps found: minor (Gemini models have no google/gemini provider in models.json by design)

Provider Model Counts

Provider	Models	Status
openai	208	✅ ok
anthropic	8	✅ ok
gemini	50	✅ ok
copilot	30	✅ ok (reflect endpoint)

Raw API Fields Discovered

• OpenAI: id, created, owned_by only — no context limits or pricing.
• Anthropic: id, display_name, created_at, max_input_tokens (200 K for 4.x, 1 M for 4.6+), max_tokens, rich capabilities object (batch, extended_thinking, vision, tool_use, etc.).
• Gemini: name, displayName, inputTokenLimit, outputTokenLimit, supportedGenerationMethods. Token limits are a reliable pricing-tier proxy.
• Copilot (reflect): flat list of model IDs only — no billing multipliers or capability flags exposed.

models.json Pricing Analysis

Missing from models.json (notable)

Model ID	Provider	Notes
gpt-4-o-preview	copilot	Legacy — in reflect but not needed.
gemini-3-pro-preview, gemini-3.1-flash-lite, gemini-3.1-flash-image, gemini-3-pro-image, gemma-4-*, imagen-4.0-*	google	No google/gemini provider section in models.json by design. All covered by aliases.
gpt-4o-search-preview, gpt-4o-mini-search-preview	openai	No dedicated alias (only caught by any).

Excluded per policy: gpt-4.1, gpt-4o, gpt-4o-mini, gpt-5.4-nano.

Historical entries retained

github-copilot: claude-sonnet-4, gemini-3.1-pro-preview (still in direct Gemini API), gemini-3.5-flash (same), gpt-5.2, gpt-5.2-codex, raptor-mini

openai: o1-mini, o1-preview, gpt-5.3-codex-spark

anthropic: claude-3-* series (15 historical entries)

Pricing discrepancies

None found. Sample validations against models-dev/api.json:

Model	Stored (input/output /token)	models.dev ($/M)	✓
gemini-2.5-pro (copilot)	1.25e-06 / 1.0e-05	1.25 / 10	✅
gpt-5.5 (copilot+openai)	5e-06 / 3e-05	5 / 30	✅
claude-sonnet-4.6 (copilot)	3e-06 / 1.5e-05	3 / 15	✅

Proposed Alias Updates

Add search alias

What: New search alias for web-search-augmented models.

Why: gpt-5-search-api, gpt-4o-search-preview, and gpt-4o-mini-search-preview are all live in the OpenAI API but only reachable via the catch-all any alias. Workflow authors need a targeted semantic alias for internet-grounded responses. gpt-5-search-api is already in models.json with pricing ($1.25/$10 per M).

{
  "search": [
    "copilot/gpt-5*search*",
    "openai/gpt-5*search*",
    "openai/gpt-4o*search*"
  ]
}

All other new model families are fully covered by existing aliases: gemini-flash-lite covers gemini-3.1-flash-lite; gemini-3.1-flash covers all 3.1-flash variants; gemini-3-pro covers gemini-3-pro-preview; gemma covers gemma-4-*; image-generation covers gemini-3.x-*-image and imagen-4.0-*; antigravity, nano-banana, robotics, deep-research cover their respective new models.

Full Model Lists by Provider

Copilot (30): claude-haiku-4.5, claude-opus-4.5/4.6/4.7/4.8, claude-sonnet-4.5/4.6, gemini-2.5-pro, gemini-3-flash-preview, gpt-4.1†, gpt-4o†, gpt-4o-mini†, gpt-5-mini, gpt-5.3-codex, gpt-5.4, gpt-5.4-mini, gpt-5.5, legacy gpt-3.5-turbo/gpt-4 variants, embedding models (†=deprecated)

Anthropic (8): claude-haiku-4-5-20251001, claude-opus-4-1-20250805, claude-opus-4-5-20251101, claude-opus-4-6/4-7/4-8, claude-sonnet-4-5-20250929, claude-sonnet-4-6

Gemini (50): gemini-2.0-flash/lite, gemini-2.5-flash/flash-image/flash-lite/pro, gemini-3-flash-preview, gemini-3-pro-image/preview, gemini-3.1-flash-image/lite/tts/pro variants, gemini-3.5-flash, gemma-4-26b/31b, imagen-4.0-fast/standard/ultra, antigravity-preview-05-2026, nano-banana-pro-preview, deep-research variants, veo/lyria/robotics/embedding models

OpenAI (208 — key production models): gpt-5 family (gpt-5, gpt-5.1–5.5 + mini/nano/pro/codex variants), gpt-5-search-api, gpt-image-1/2, gpt-4o-search-preview, gpt-4o-mini-search-preview, o1/o1-pro, o3/o3-mini/o3-pro/o3-deep-research, o4-mini/o4-mini-deep-research, computer-use-preview (+ 173 experimental/versioned/audio/embedding entries)

Notes

• make refresh-models-json failed (403). Re-run when models.dev catalog recovers.
• raptor-mini, claude-sonnet-4, gemini-3.1-pro-preview, gemini-3.5-flash retained as historical entries in github-copilot provider.

Warning

Firewall blocked 1 domain

The following domain was blocked by the firewall during workflow execution:

• models.dev

To allow these domains, add them to the network.allowed list in your workflow frontmatter:

network:
  allowed:
    - defaults
    - "models.dev"

See Network Configuration for more information.

Generated by 📦 Daily Model Inventory Checker · 160 AIC · ⌖ 21.7 AIC · ⊞ 7.4K · ◷

• expires on Jun 28, 2026, 4:34 PM UTC-08:00

[github-actions]

This issue is being closed as outdated. A newer issue has been created: #40909

View newer issue

---

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