Respan's Span-02 brings anytime reasoning: interrupt it mid-thought and still get the best answer
techNmak · x · 2026-10-16
Respan launched Span-02, a decision model built around "anytime reasoning": instead of picking an effort level, you set a deadline and can interrupt it anytime to get its current best decision with confidence — it improves the longer it thinks.
- With reasoning off, it beats Jev at 1/4 the price and runs at 1/3 the latency of decision models at its level.
- Multimodal (reads images and screenshots), and can return numbers, scores, or choices rather than just probabilities.
- Configurable reasoning time and fully open weights.
The author notes this applies the classic computer science idea of anytime algorithms: agents can work within a time budget and escalate uncertain decisions instead of waiting for every reasoning process to finish.
More from coding & agent
- Tool calls are the hidden bottleneck killing AI coding speed — Crabbox and 5 fixes — steipete · 2026-10-16
- Agentic sandbox WorldBox doubles GPT-6 Astra's MazeBench score, 23 to 61 gems — ZhitingHu · 2026-10-16
- Two prompts: Gemini 4 Argon turns a single-player game into full multiplayer — chongdashu · 2026-10-16
- Anthropic and Gray Swan launch $300,000 indirect prompt injection red-teaming contest — bcherny · 2026-10-16
- Vercel launches Token Pass: $30/mo for $80 of GLM 5.3 and Kimi K3 credits — evilrabbit_ · 2026-10-16
- Muse Spark Open goes live on Together AI as a Day 0 launch partner — togethercompute · 2026-10-16