Repository navigation
feat: add NEAR AI pricing and model configs - #982
Conversation
Adds pricing/near-ai.json and general/near-ai.json for NEAR AI Cloud (https://near.ai), an OpenAI-compatible private inference provider. 53 models (52 chat, 1 embedding) from NEAR's public catalog (GET https://cloud-api.near.ai/v1/model/list). NEAR reports prices as amount x 10^-scale USD per token; converted to cents per token (amount x 10^-7), e.g. claude-sonnet-4-5 3000 -> 0.0003. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
🔵 Needs a closer look
The large externally sourced pricing and capability dataset requires human verification against NEAR’s live catalog.
0 open findings
What changed in this PR
Adds NEAR AI model metadata and token pricing for the provider’s catalog.
Changes:
- Configures 52 chat models and one embedding model.
- Adds per-model input and output token pricing.
| File | Description |
|---|---|
general/near-ai.json |
Defines model capabilities and parameters. |
pricing/near-ai.json |
Defines token rates and cost calculations. |
🧠 Review effort: Balanced
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
|
@narengogi can we add cache read and write pricing. |
|
@narengogi this PR looks good, we can also add cache_read pricing and add the openai/privacy-filter model if we want to support it for this provider. There are individual models under decisions, image, rerank and transcription categories which are not in this PR, can be added for complete support of 58 models from this provider. |
Cache read for the 47 models NEAR prices it for, cache write for the 7 that publish it (textPricing default tier). Long-context and priority/flex tiers are not modelled. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
changes made |
pricing/near-ai.jsonandgeneral/near-ai.json(53 models: 52 chat, 1 embedding), priced from NEAR's public catalog🤖 Generated with Claude Code