Skip to content

Search stories, models, companies…Subscribe

Topic · Concept

Open models

Open models are AI models whose weights are published for anyone to download and run, usually under a license that permits some reuse. The term covers a wide range, from fully open-source releases to "open-weight" models with restrictive licenses, and the difference matters for what you can legally do with them.

50 stories · updated Oct 26

Follow via RSS Get Open models alerts

Key facts

Checked Oct 8

Open Source AI Definition: Version 1.0, published October 28, 2024 (OSI)
Required components (OSAID): Data information, code, parameters
Four freedoms (OSAID): Use, study, modify, share, for any purpose

Latest on Open models

Models · NewsKimi K3.1 Pro: $2/$10 per million tokens, a 55 on Artificial Analysis, and open weights promised by November 4Kimi K3.1 Pro, released by Moonshot AI on October 23, 2026, costs $2 input and $10 output per million tokens, scores 55 on the Artificial Analysis Intelligence Index, and is due as open weights by November 4. Its price, benchmarks and what isn't known yet.Oct 26·5 min read

Industry · NewsDeepSeek's $15 billion funding round: Bloomberg and Reuters report the 100 billion yuan raise, days after the US federal ban was made publicDeepSeek's $15 billion funding round: Bloomberg reported on October 22, 2026 that DeepSeek has raised about 100 billion yuan, and Reuters reported the round was set to complete within days at about 500 billion yuan before the new money. It's the round 9to5AI first reported on October 16. DeepSeek hasn't confirmed it.Oct 24·4 min read

Models · NewsMistral Large 4: $1.36/$4.18 per million tokens, a 1M-token context claim, mid-pack independent scores, and open weights due October 31Mistral Large 4 is a roughly 1-trillion-parameter MoE with 52B active parameters, in API preview since October 6, 2026 at $1.36 input and $4.18 output per million tokens. Independent evaluators place it mid-pack, and the open weights are scheduled for October 31.Oct 24·5 min read

Research · BenchmarksQwen3-Coder writes less secure code for a US defense contractor, Hirundo says: what the study found, and what it doesn't proveIn an October 21, 2026 study, Hirundo says Alibaba's Qwen3-Coder-30B-A3B wrote more hackable web apps when told the code was for a US defense contractor than for a Chinese one. Here's how the test worked, the numbers, the statistics, the limits, and how it compares with CrowdStrike's DeepSeek study.Oct 24·8 min read

Open source · NewsDeepSeek DeepInfer: the Apache 2.0 inference engine behind V4.1 Flash runs on Nvidia Blackwell and Huawei Ascend, with key serving parts held backDeepSeek DeepInfer is the inference engine DeepSeek says serves DeepSeek-V4.1-Flash in production, open-sourced under Apache 2.0 on October 20, 2026. It supports one model on two kinds of hardware, and the KV-cache and load-balancing systems aren't included.Oct 24·5 min read

Models · NewsVoxtral 3: Mistral's new speech-to-text, text-to-speech and voice-agent SDK are API-only, unlike earlier Voxtral modelsVoxtral 3, released October 20, 2026, adds three Mistral transcription models, a text-to-speech model and a voice-agent SDK priced from $0.0035 a minute. Unlike earlier Voxtral releases, none of the Voxtral 3 models has open weights.Oct 24·5 min read

Open source · NewsRecursive Flint: an Apache 2.0, 40B coding model with 2.5B active parameters and a self-reported 89.7% on SWE-bench Pro v2Recursive Flint is a 40B-parameter mixture-of-experts coding model with about 2.5B active parameters, released under Apache 2.0 on October 21, 2026. Its 89.7% on SWE-bench Pro v2 is Recursive's own number and wasn't on Scale AI's leaderboard as of October 22.Oct 24·5 min read

Products · NewsPerplexity Computer on Mac: what it does locally, who can get it, and the permissions it asks forPerplexity Computer on Mac can now run fully local: since October 20, 2026, the model, the agent and the task run on an Apple silicon Mac for Pro, Max and Enterprise subscribers. What 'local' covers, the 36GB memory recommendation, and the macOS permissions it asks for.Oct 24·5 min read

Policy · NewsDeepSeek banned in US federal agencies: what the DHS order requires, and where else DeepSeek is bannedDeepSeek is now banned in US federal civilian agencies. A Department of Homeland Security order, made public on October 19, 2026, bars agencies from buying DeepSeek products and requires them to remove existing ones, including third-party tools built on DeepSeek. What the order says, what it leaves out, and a dated list of the other governments that ban or restrict DeepSeek.Oct 20·7 min read

Open source · NewsMistral Timestral: four open-weight time-series models, a $20M revenue cap, and a CERN statement of intentMistral Timestral is a family of four time-series forecasting models, from 5M to 209M parameters, with weights on Hugging Face under a revenue-capped 'Modified MIT' license. The same week, CERN and Mistral signed a statement of intent, not a contract, to build a Large Physics Model.Oct 20·5 min read

Nothing in this format on this page — try another filter or the next page.