Baltor Get started

Endpoints

CoreWeave

A hosted service with its own key and prices. This page lists the addresses it answers, the facts its documentation states, and the setup of each harness.

Addresses and facts

  • OpenAI Chat Completions: https://api.inference.wandb.ai/v1 models.dev, read
Authentication
Authorization: Bearer, with the key in WANDB_API_KEY models.dev, read
Structured output
Unknown
Tool calling
29 of 29 listed models The share of this provider's models that models.dev records with tool calling. models.dev, read
Rate limits
Unknown
Prices
Unknown
Data retention
Unknown
Documentation
Read the page models.dev, read
Output limits Baltor recorded
Unknown

Harness setup

Replace JetBrains/Mellum2-12B-A2.5B-Instruct with the model you want.

OpenCode

Put this in opencode.json in your project folder:

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "wandb": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "CoreWeave",
      "options": {
        "baseURL": "https://api.inference.wandb.ai/v1",
        "apiKey": "{env:WANDB_API_KEY}"
      },
      "models": {
        "JetBrains/Mellum2-12B-A2.5B-Instruct": {
          "name": "JetBrains/Mellum2-12B-A2.5B-Instruct"
        }
      }
    }
  }
}
  • OpenCode reads any OpenAI-compatible address through the @ai-sdk/openai-compatible package, and an address that speaks the Responses API through @ai-sdk/openai.

From OpenCode documentation, read .

Pi

Put this in ~/.pi/agent/models.json:

{
  "providers": {
    "wandb": {
      "baseUrl": "https://api.inference.wandb.ai/v1",
      "api": "openai-completions",
      "apiKey": "$WANDB_API_KEY",
      "models": [
        {
          "id": "JetBrains/Mellum2-12B-A2.5B-Instruct"
        }
      ]
    }
  }
}
  • The apiKey field can name an environment variable as $NAME.

From Pi documentation, read .

Codex

Codex speaks only the Responses API, and CoreWeave documents no Responses address. A gateway that offers one can sit in between.

  • Codex speaks the Responses API only: responses is the one supported wire API of a custom provider. Ollama and LM Studio are built in and start with --oss.

From Codex documentation, read .

Claude Code

Claude Code sends Anthropic Messages requests, and CoreWeave documents no such address. A gateway that translates to that API can sit in between.

  • Claude Code sends Anthropic Messages requests to ANTHROPIC_BASE_URL. Anthropic says it does not support routing Claude Code to models other than Claude through any gateway, so some features may not work with another model.

From Claude Code documentation, read .

Models it lists

Prices in US dollars per million tokens, input and output, as models.dev, read records them.

ModelInputOutputContextAs of
Mellum2 12B A2.5B JetBrains/Mellum2-12B-A2.5B-Instruct0.050 USD0.100 USD131,072 older than 30 days
MiniMax M3 MiniMaxAI/MiniMax-M30.230 USD0.960 USD262,144 older than 30 days
Qwen3 14B Instruct OpenPipe/Qwen3-14B-Instruct0.050 USD0.220 USD32,768 older than 30 days
Qwen3 30B A3B Instruct 2507 Qwen/Qwen3-30B-A3B-Instruct-25070.100 USD0.300 USD262,144 older than 30 days
Qwen3.5-35B-A3B Qwen/Qwen3.5-35B-A3B0.250 USD1.25 USD262,144 older than 30 days
Qwen3.6 27B Qwen/Qwen3.6-27B0.600 USD3.60 USD262,144 older than 30 days
Qwen3.6 35B A3B Qwen/Qwen3.6-35B-A3B0.250 USD1.25 USD262,144 older than 30 days
Qwen3.8 27B Qwen/Qwen3.8-27B0.400 USD3.00 USD262,144 older than 30 days
DeepSeek V3.1 deepseek-ai/DeepSeek-V3.10.550 USD1.65 USD161,000 older than 30 days
DeepSeek V4 Flash deepseek-ai/DeepSeek-V4-Flash0.140 USD0.280 USD1,048,576 older than 30 days
DeepSeek V4 Flash 0731 deepseek-ai/DeepSeek-V4-Flash-07310.130 USD0.280 USD262,144 older than 30 days
DeepSeek V4 Pro deepseek-ai/DeepSeek-V4-Pro1.15 USD2.55 USD1,048,576 older than 30 days
DeepSeek V4 Pro 0813 deepseek-ai/DeepSeek-V4-Pro-08131.31 USD3.96 USD1,048,576 older than 30 days
DeepSeek V4.1 Flash deepseek-ai/DeepSeek-V4.1-Flash0.200 USD0.650 USD1,048,576
Gemma 4 26B A4B google/gemma-4-26B-A4B-it0.100 USD0.300 USD262,144 older than 30 days
Gemma 4 31B google/gemma-4-31B-it0.100 USD0.340 USD262,144 older than 30 days
Granite 4.1 8B ibm-granite/granite-4.1-8b0.050 USD0.100 USD131,072 older than 30 days
Granite 4.2 8B ibm-granite/granite-4.2-8b0.100 USD0.150 USD131,072 older than 30 days
Llama 3.1 70B meta-llama/Llama-3.1-70B-Instruct0.800 USD0.800 USD131,072 older than 30 days
Llama 3.1 8B meta-llama/Llama-3.1-8B-Instruct0.220 USD0.220 USD131,072 older than 30 days
Llama 3.3 70B meta-llama/Llama-3.3-70B-Instruct0.710 USD0.710 USD128,000 older than 30 days
Kimi K2.6 moonshotai/Kimi-K2.60.650 USD3.41 USD262,144 older than 30 days
Kimi K2.7 Code moonshotai/Kimi-K2.7-Code0.710 USD3.50 USD262,144 older than 30 days
Nemotron 3 Ultra nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B0.500 USD2.15 USD262,144 older than 30 days
Nemotron 3.5 Lightning nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B0.070 USD0.200 USD262,144 older than 30 days
gpt-oss-120b openai/gpt-oss-120b0.030 USD0.170 USD131,072 older than 30 days
gpt-oss-20b openai/gpt-oss-20b0.030 USD0.130 USD131,072 older than 30 days
GLM 5.2 zai-org/GLM-5.20.760 USD2.42 USD1,048,576 older than 30 days
GLM 5.3 Flash zai-org/GLM-5.3-Flash0.150 USD0.500 USD1,048,576

Sources of this page