Span-02
- Family
- Span
- Released
- 2026-10-15
- Context
- 128K
- Parameters
- 26B
- Architecture
- Mixture of Experts
- Openness
- Open source
- License
- Apache 2.0OSI-approvedPermitted
- Weights
- Available
- Code
- Unknown
- Training
- Fine-tuned
Cheapest of 1 route · OpenRouter
About
Span-02 is Respan's open-weight decision model, a fine-tune of Google's Gemma 4 26B-A4B with about 4B active parameters. Given text and up to four images, it returns a typed answer: a choice, yes/no/unknown, a score on a scale, or a number. It uses anytime reasoning: reasoning can be off, set to a time budget after which it returns its best decision so far, or chosen per request in auto mode. Context is 131,072 tokens per question. Weights ship in NVFP4/FP8 with Respan's own Rust/CUDA engine for one NVIDIA Blackwell GPU; Respan also serves it through OpenRouter.
Providers(1)
Compare API pricing across 1 tracked provider route for input and output tokens, batch, and cached reads when available.
| Provider | Input / 1M | Output / 1M | Route |
|---|---|---|---|
| OpenRouter | $0.010 | Free | Serverless |
Capabilities
Benchmark peer barsfor RAG
No task-mapped benchmark peers are available for this model yet.
Migration checks
No linked migration route is available for this model yet.
API versions
respan/span-02Span 2Rankings & picks(1)
Cheapest of 1 route · OpenRouter