modelgrep

Best Nous Models for Coding

Quick answer · Updated June 2026

Hermes 3 405B Instruct (free) is the best Nous model for coding, with a 18.1 Artificial Analysis Coding Index across benchmarks like SWE-bench and SciCode. Hermes 3 405B Instruct (18.1) and Hermes 4 405B (16.0) round out the top three.

18.1Coding
17.6Intelligence
22 t/sSpeed
FreeInput /M
131KContext

AI models ranked by the Artificial Analysis Coding Index, measuring real-world software engineering ability across benchmarks like SWE-bench, SciCode and terminal tasks. The best LLMs for code generation, debugging and agentic development.

  1. 1N
    hermes-3-llama-3.1-405b:free
    17.6 intel · Free/M · 22 t/s
    18.1
    Coding
  2. 2N
    hermes-3-llama-3.1-405b
    JSON17.6 intel · $1.00/M · 22 t/s
    18.1
    Coding
  3. 3N
    hermes-4-405b
    ReasoningJSON18.6 intel · $1.00/M · 31 t/s
    16.0
    Coding
  4. 4N
    hermes-4-70b
    ReasoningJSON16.0 intel · $0.130/M · 81 t/s
    14.4
    Coding
  5. 5N
    hermes-3-llama-3.1-70b
    JSON12.6 intel · $0.700/M · 21 t/s
    9.2
    Coding

Frequently asked

What is the best Nous model for coding?

Hermes 3 405B Instruct (free) is the best Nous model for coding, with a 18.1 Artificial Analysis Coding Index across benchmarks like SWE-bench and SciCode. Hermes 3 405B Instruct (18.1) and Hermes 4 405B (16.0) round out the top three.

What's a good alternative to Hermes 3 405B Instruct (free)?

Hermes 3 405B Instruct (18.1) is the closest alternative on this metric, followed by Hermes 4 405B (16.0). See the full ranking above for the tradeoffs.

How many Nous models are there?

modelgrep tracks 5 Nous models with live benchmarks, speed, latency and per-provider pricing, led on intelligence by Hermes 4 405B. 5 of them qualify for this ranking.

More Nous rankings

All rankings