Models / Nous
Hermes 4 70B
by Nous ·
nousresearch/hermes-4-70bHermes 4 70B is a hybrid reasoning model from Nous Research, built on Meta-Llama-3.1-70B. It introduces the same hybrid mode as the larger 405B release, allowing the model to either...
reasoningreleased 2025-08-26
Legal score · raw
-
not yet benchmarked
Context
131K
max output 118K
Input
$0.13
per 1M tokens
Output
$0.40
per 1M tokens
Suite cost
-
run the suite to see
Benchmark results
| Task | Category | Raw | Juiced | Correct | Latency | Cost | Ran |
|---|---|---|---|---|---|---|---|
| Hearsay Identification | Evidence | - | - | - | - | - | - |
| Bluebook Citation Format | Research & Writing | - | - | - | - | - | - |
| Federal Civil Procedure | Procedure | - | - | - | - | - | - |
| Limitations Arithmetic | Procedure | - | - | - | - | - | - |
| Contract Clause Classification | Contracts | - | - | - | - | - | - |
| Citation Hallucination Resistance | Reliability | - | - | - | - | - | - |
Measured by DocketBuster
These numbers come from DocketBuster's own legal battery, not from DocketRouter's suite. Latest run per metric, with n and a 95% Wilson interval where the source reports one. See docketbuster.com/benchmarks.
| Metric | Value | n | Interval | Measured |
|---|---|---|---|---|
| Statute pinpoint, exact section (no retrieval) | 12.4% (37/299) | 299 | 95% CI 9.1% to 16.6% | 2026-08-23 |
| Statute pinpoint, exact section (with DocketBuster retrieval) | 76.2% (234/307) | 307 | 95% CI 71.2% to 80.6% | 2026-08-24 |
| Say-nothing rate (declines to bluff when the answer is not in the record) | 96.3% | 300 | 95% CI 93.6% to 97.9% | 2026-08-23 |
| Abstained on statute pinpoint | 9.8% (30/307) | 307 | count, no interval reported | 2026-08-24 |
| Coaching quality (GW-14x, 0 to 8) | 6.60 / 8 | - | rubric mean, no interval reported | 2026-08-24 |
Source files: hard-llm-ortier-hermes4-70b.json, hard-llm-level-hermes4-70b-statute_rag.json, gw14x-ortier-hermes4-70b.json. Raw model name in source: nousresearch/hermes-4-70b.