InternLM3
Shanghai AI Lab's 8B open model — strong reasoning in a small package, trained on 4T tokens.
About
InternLM3 is the third generation of Shanghai AI Lab's InternLM series, an 8-billion-parameter open-weights model released in January 2025. It was trained on 4 trillion tokens — far fewer than peers at its size — and ships with a thinking mode for step-by-step reasoning. The lab reports competitive scores against larger open models on reasoning and coding benchmarks, though independent third-party evaluations are thinner than for the biggest Western releases. Weights are Apache 2.0 licensed.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Team-reported: competitive with larger open models on reasoning and coding benchmarks (see technical report); independent evals limited.
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Official InternLM GitHub lists 2025-01-15 release for the 8B instruct; Apache-2.0 per curated open-weight lists; Thinking mode for step-by-step reasoning. License: Apache-2.0..
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 7.0/10
- Maintenance · recency, releases, issue load
- 8.0/10
- Openness · license + docs
- 9.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · Official InternLM GitHub lists 2025-01-15 release for the 8B instruct
- · Apache-2.0 per curated open-weight lists
- · Thinking mode for step-by-step reasoning