gpt-oss-120b alternatives

verdictfirst on capability and first on licence at the same time, which almost never happens — this is the default open-weight choice right now.

switch foranyone who wants the strongest open model available and needs the legal answer to be boring.

verdictarguably the best coding model you can legally do anything with — carrying a 'preview' label that deepseek has not removed on the schedule it set itself.

switch forcoding and reasoning workloads where you want frontier results without a frontier bill or a licence review.

verdictthe highest-ranked open model on lmarena after gemini, licensed apache 2.0 with nothing bolted on — the main catch is that it's now eight months old.

switch forteams who want top-tier open coding performance from a european vendor with unambiguous terms.

verdictexcellent at the job it was built for, with a licence quirk almost nobody will hit — but which press coverage summarising it as 'mit' never mentions.

switch forlong-running agentic engineering work, for anyone comfortably below the attribution thresholds.

verdictthe strongest open-weight model from a us lab, built for customisation rather than benchmarks — and its own model card says plainly that it isn't the strongest overall.

switch forteams who want a permissive multimodal base to fine-tune rather than a leaderboard winner to deploy.

verdictthe best price-to-capability ratio in the category, and widely mistaken for a variant of v4 pro when it's its own release with its own weights.

switch forhigh-volume work that doesn't need pro-tier reasoning but does need a licence you can ignore.

verdictthree generations of custom gemma terms replaced by plain apache 2.0, on a family built to run on hardware you already own — the caveats are housekeeping rather than substance.

switch foron-device and laptop-class multimodal work where the model has to run locally and legally.

verdictan unusually strong small model under clean apache 2.0 — the trap is definitional, because the 'qwen' name also covers a closed api-only flagship.

switch foragentic coding on modest hardware, for teams who want alibaba's engineering without alibaba's api.

verdictcapable, long-context and well-engineered — marked down purely because we could not establish, from nvidia's own sources, what you are agreeing to.

switch forenterprises with legal teams who can get nvidia to say which licence actually governs before deployment.

verdictevery other entry calls itself open and means the weights. ai2 publishes the training data too — which is worth more than the benchmark places it gives up.

switch forresearch, auditing, and anyone who needs to know what a model was actually trained on.