Nimble lands in Ollama with typed decisions

// 1h agoMODEL RELEASE

Nimble lands in Ollama with typed decisions

Bespoke Labs’ Apache 2.0 Nimble is a 9B Qwen3.5-9B fine-tune for fast local classification, routing, moderation, and scoring. Ollama now exposes it through the /v1/systemone API, where it reports 75.7% accuracy across 13 public datasets—just behind Jev’s 76.0%.

// ANALYSIS

Nimble makes a compelling case for replacing unnecessary text generation with bounded, probability-backed decisions, though its benchmark parity with Jev is highly task-dependent.

–Local inference reduces latency, cost, and data exposure for routing and moderation workloads.
–Directly scoring allowed answer tokens produces typed outputs without parsing generated JSON.
–Nimble leads Jev on some tasks, but trails sharply on moderation and German-language routing.
–Its 9B footprint and flat-schema limitations make it less flexible than a general-purpose LLM.
–The reported results are promising, but developers should benchmark their own policies before automating high-impact decisions.

// TAGS

nimblellmsmall-llmopen-weightsopen-sourcelocal-firstinferencestructured-output

DISCOVERED

1h ago

2026-09-30

PUBLISHED

1h ago

2026-09-30

RELEVANCE

9/ 10

AUTHOR

DIY Smart Code

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

UPDATE1h ago

0xDesigner Doubles Capacity, Adds Fast Mode

0xDesigner MCP lets coding agents hand UI tasks to a human designer. After reaching weekly capacity less than 24 hours after launch, it now offers Normal 48-hour and Fast 24-hour delivery, with Fast costing 1.5× credits.

UPDATE1h ago

Ollama 0.35 Adds Local Decision Models

Ollama 0.35 adds the `/v1/systemone` API for local decision models that return typed choices, probabilities, and scores instead of prose. The feature supports ticket triage, model routing, policy checks, and content classification through models including Nimble and Tev1.

TUTORIAL1h ago

Aura Makes DESIGN.md Design Memory

This tutorial shows how Aura uses reusable DESIGN.md files to carry typography, color, spacing, layout, motion, and interaction rules across AI-generated websites. Aura also supports URL-based design extraction, visual editing, and HTML export.