Skip to content

docs(readme): resync model tables with live catalog; fix market data, Surf, Live Search, prices - #114

Merged
VickyXAI merged 1 commit into
mainfrom
docs/readme-catalog-resync
Nov 15, 2026
Merged

VickyXAI merged 1 commit into
mainfrom
docs/readme-catalog-resync

Conversation

@VickyXAI

Copy link
Copy Markdown
Contributor

README + CHANGELOG (Unreleased → Docs) only. No code changes.

Every change was checked against the live gateway on 2026-11-15 (GET https://blockrun.ai/api/v1/models, plus unpaid 402 / status probes).

Model tables

  • Added (all available: true in the catalog, context from context_window): anthropic/claude-fable-5.2 (1M), google/gemini-4-carbon (1M), moonshot/kimi-k3.1-pro (1M, replaces kimi-k3, hidden but callable), deepseek/deepseek-v4.1-flash (1M), openai/gpt-6.1-sol (1.05M), qwen/qwen3.8-max (1M), mistral/mistral-large-4 (512K), tencent/hy4-preview (1M), poolside/laguna-s-2.1 (1M), free nvidia/minimax-m3.1-flash (1M).
  • Removed: openai/gpt-4.1-nano (retired, gone from catalog), openai/dall-e-3 (not in the catalog's image category).
  • Image table: added the image-category models the catalog lists (gpt-image-2.5-flare, gpt-image-2.5-sunburst, nano-banana-2, grok-imagine-image-2.0, seedream-5-pro).

Free tier

  • Pin-worthy list and the free routing-profile row now match FREE_TIERS in router_adapter.py.
  • New warning not to pin nvidia/llama-3.2-11b-vision, nvidia/nemotron-3-super-120b, cohere/north-mini-code, nvidia/muse-glimmer-30b (they answer 200 but are served by another model).

Market data (PriceClient)

  • Not Pyth, not realtime: crypto / FX / commodity price, history and lists are free daily closes (source: chainlink|backpack, realtime: false).
  • Stock / usstock price and history answer 501 on blockrun.ai and sol.blockrun.ai, and stock lists still answer 200 for all 12 markets. PriceClient raises the 501 as APIError(status_code=501), API error: 501: US Stock price is not available (checked live).

Surf: every /v1/surf/* probe answers 410 endpoint_retired on both gateways. The section now documents the deprecation (matching SurfClient's DeprecationWarning) and points to the replacements.

Live Search: the example used openai/gpt-5.2, and the SDK sends search_parameters, which the gateway rejects with 400 SEARCH_UNSUPPORTED (xai/* only). The example now uses xai/grok-4.7.

Prices: dropped numbers the 402 quotes contradict (Live Search $0.26/$0.13, SearchClient per-source pricing, Predexon $0.001/$0.005, sound effects $0.05, DeepSeek V4 Pro $0.435/$0.87, FAQ "$0.10/M"). Also dropped the testnet prices, because testnet.blockrun.ai was unreachable. Modal pricing now matches the gateway ($0.01 CPU, $0.05–$0.40 GPU, hourly when longer). The FAQ no longer says "no API keys".

Checks: pytest tests/unit 1569 passed; node scripts/sync-brand-numbers.mjs --check up to date. The br: markers were left untouched.

…f, Live Search, prices

- Add catalog models missing from the tables (claude-fable-5.2, gemini-4-carbon,
  kimi-k3.1-pro, deepseek-v4.1-flash, gpt-6.1-sol, qwen3.8-max, mistral-large-4,
  hy4-preview, laguna-s-2.1, free minimax-m3.1-flash) and the new image models;
  drop retired gpt-4.1-nano and dall-e-3.
- Free tier: list the pin-worthy free models, warn not to pin the four ids the
  gateway now serves with another model, fix the free routing-profile row.
- PriceClient: free daily closes (Chainlink/Backpack), not Pyth or realtime;
  equity price/history answer 501, raised as APIError(status_code=501).
- Surf: /v1/surf/* is retired (410), SurfClient deprecated.
- Live Search example used openai/gpt-5.2 (400 SEARCH_UNSUPPORTED); use xai/grok-4.7.
- Drop per-call prices the live 402 quotes no longer match; fix Modal pricing.
@VickyXAI
VickyXAI merged commit 58c4064 into main Nov 15, 2026
6 checks passed
@VickyXAI
VickyXAI deleted the docs/readme-catalog-resync branch November 15, 2026 18:48
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant