Model comparison
Codestral vs Phi-2
Codestral has enough public results to be ranked (#290); Phi-2 does not yet, so treat this comparison as directional.
Last verified . 2 shared benchmarks.
Mistral AI
MicrosoftSummary
- They share 2 benchmarks with published results for both.
- Phi-2 has downloadable open weights; the other is API-only.
Side by side
| Codestral | Phi-2 | |
|---|---|---|
| Provider | Mistral AI | Microsoft |
| Noometry Index | 30.6 | — |
| Released | 2024-05-29 | 2023-12-12 |
| Weights | Proprietary | Open |
| Context window | 256K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.30 | — |
| Output $ / M tokens | $0.90 | — |
| Results tracked | 7 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Codestral: 27.3 (#321), Phi-2: —
| Benchmark | Codestral | Phi-2 |
|---|---|---|
| HumanEval+ | 73.8% | 45.1% |
| MBPP+ | 61.9% | 54.2% |
| Aider Polyglot | 11.1% | — |
| BigCodeBench Instruct | 41.8% | — |
| BigCodeBench Complete | 52.5% | — |
| ALE-Bench | 137.78 | — |
Reasoning Not comparable
Codestral: 19.8 (#251), Phi-2: —
| Benchmark | Codestral | Phi-2 |
|---|---|---|
| Kagi LLM Benchmark | 32.5% | — |
| Adversarial NLI | — | 42.5% |
| BIG-Bench Hard | — | 59.4% |
| Epoch Capabilities Index | — | 107.94 |
| HellaSwag | — | 53.6% |
| WinoGrande | — | 54.7% |
Knowledge Not comparable
Codestral: —, Phi-2: —
| Benchmark | Codestral | Phi-2 |
|---|---|---|
| ARC (AI2) Challenge | — | 75.9% |
| MMLU | — | 58.4% |
| OpenBookQA | — | 73.6% |
| TriviaQA | — | 45.2% |
Frequently asked questions
Is Codestral better than Phi-2?
Codestral has enough public results to be ranked (#290); Phi-2 does not yet, so treat this comparison as directional.
How many benchmarks do Codestral and Phi-2 share?
2 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and Phi-2 has 11.