InternLM
Shanghai AI Lab · China · Language & multimodal
China Open weightsFree tier
Showing version: InternLM3 (InternLM3-8B-Instruct) · 2025-01-15 · Version lineage ↓
01
Overview
InternLM (Shusheng Puyu) is Shanghai AI Laboratory's open LLM series, evolving through InternLM, InternLM2 and InternLM2.5 to InternLM3, known for open weights, low training cost and a deep-thinking mode.
Release date2025-01-15
02 / Provenance
Where each field comes from
On this entry, 7 fields have no official page to link to directly. Their values come from official or institutional pages that open and can be quoted verbatim, cited next to the number and under Practical info; anything without a citation stays blank rather than estimated.
03 / Plain layer
Plain-language benchmarks
Math 83%0-shot
Math 20%0-shot
Reasoning 37.4%0-shot
Chinese 83.1%0-shot
Coding 17.8%2407-2409, Pass@1
Memory Not disclosed
Model size Not disclosed
Price / free Not disclosed开源模型, 权重可从 GitHub/Hugging Face/ModelScope 免费下载; 无官方按量 API 定价
04
Practical facts
| Context length | Not disclosed |
|---|---|
| Max output | Not disclosed |
| Parameters | Not disclosed |
| Input | Not disclosed |
| Output price | Not disclosed |
| CNY | Not disclosed |
| Free tier | Open-source model; weights can be downloaded free from GitHub, Hugging Face and ModelScope; no official pay-as-you-go API pricing. |
| Output speed | Not disclosed |
| Release date | |
| License | Not disclosed |
Show technical data
Technical benchmark table
| Benchmark | Score | Max / unit | Source type | Source | Test date | Conditions |
|---|---|---|---|---|---|---|
| CMMLU cmmlu | 83.1% | percent / 100 | Vendor-reported | 0-shot | ||
| MMLU-Pro mmlu-pro | 57.6% | percent / 100 | Vendor-reported | 0-shot | ||
| GPQA Diamond gpqa-diamond | 37.4% | percent / 100 | Vendor-reported | 0-shot | ||
| MATH-500 math-500 | 83% | percent / 100 | Vendor-reported | 0-shot | ||
| AIME aime · 2024 | 20% | percent / 100 | Vendor-reported | 0-shot | ||
| LiveCodeBench livecodebench | 17.8% | percent / 100 | Vendor-reported | 2407-2409, Pass@1 | ||
| IFEval ifeval | 79.3% | percent / 100 | Vendor-reported | Prompt-Strict |
| Version | Release date | Status | Openness | Parameters | What changed | Detail | |
|---|---|---|---|---|---|---|---|
| InternLM3 (InternLM3-8B-Instruct)Showing version | Active | Open weights | Not disclosed | TBD | Detail | ||
| InternLM2.5 | Active | Open weights | Not disclosed | TBD | Detail | ||
| InternLM2 | Active | Open weights | Not disclosed | TBD | Detail | ||
| InternLM | Active | Open weights | Not disclosed | TBD | Detail |
Last updated: · Sources on detail page
06 / Compare