Gutsy 0.8B v0.3
Data: JevBench v1.6.39, published 2026-10-22
Capability 43.1 · Not ranked in the current release
latency 2.7× Jev (v1.5 reference)
Last measured: 2026-10-10 · measurement release: v1.6.5
Licence not stated
intelligence
8.6
calibration
77.5
speed
72.9
cost
78.4
Composite (secondary)
0.8 · Composite rank #157
Cost and measurement conditions
$0.0053 per 1,000 decisions (estimated)
ESTIMATE (*est.), input tokens only. The author's gutsy-inference 0.3.0 server emits no output-token field in any spelling; under the chief-of-staff decision of 9 Oct 2026 a missing output_tokens field on a native option-probability model that generates no text is treated as 0 output tokens, exactly as a server reporting a literal 0 is treated. Zero output is a verified property of this backend, not an assumption: it has no generation path at all (no generate/create_completion/n_predict/max_tokens/sampler); it prefills via llm.eval(tokens) and reads llama_get_logits_ith(ctx,-1) (r52 PREREG-AMENDMENT-4). Rate: the declared base Qwen/Qwen3.5-0.8B resolves in the frozen 25 Sep 2026 snapshot to deepinfra:Qwen/Qwen3.5-0.8B at USD 0.01 per 1M input and USD 0.05 per 1M output. This row's own measured usage on draw v1.6-regular-20261010-a2: 789,980 input tokens and 0 output over 1,500 decisions => USD 0.0052665333333333335 per 1,000 decisions.
Measured p50 latency: 0.756 s (on-GPU, no network, loopback; not directly comparable with API latency, which we time end to end from our measuring host, network included)
Ranking and cap use: 1.66 s (est. incl. network (2x + 150 ms))
Self-hosted model: measured p50 0.756 s on-GPU, no network (timed on our GPU over loopback). Not directly comparable with API latency, which we time end to end from our measuring host, network included. The ranking and latency cap use the same estimate for every self-hosted row: 2 × measured + 150 ms = 1.66 s. Model authors can ask us to time a hosted endpoint instead (on request).
x2 + 0.15 s (assumption, not measured)
evaluator-owned Lium pod (2xRTX3090 node, Germany), CPU-only by the author's own config (n_gpu_layers 0), 32 vCPU, egress cut with iptables and proven before the sealed input was uploaded; the author's own gutsy-inference 0.3.0 with their models.example.json settings unchanged (n_ctx 8192, reject_slot true, n_threads null)