Open-weight hybrid model for enterprise chat, coding, retrieval-augmented generation, and tool-calling workloads
1 provider
Every model in the catalog with source-linked pricing, context limits, provider availability, and published benchmark results.
Open-weight hybrid model for enterprise chat, coding, retrieval-augmented generation, and tool-calling workloads
gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized...
Efficient multimodal model for instruction following, coding, reasoning, and function calling
Popular open Llama workhorse for multilingual chat, coding, and self-hosting
| Model | Creator | Inputs | Context | Input | Output | Released | Compare |
|---|---|---|---|---|---|---|---|
| Granite-4.0-H-Smallibm/granite-4-h-small | 131.072K | $0.064 | $0.265 | 2025-10-02 | |||
| OpenAI: gpt-oss-120bopenai/gpt-oss-120b | 131.072K | $0.037 | $0.17 | 2025-08-05 | |||
| Mistral Small 3.1 24Bmistral/mistral-small-3-1-24b-instruct-2503 | 128K | $0.106 | $0.318 | 2025-03-17 | |||
| Llama-3.3-70B-Instructmeta/llama-3.3-70b-instruct | 128K | $0.1 | $0.32 | 2024-12-06 |