{"slug":"ante","name":"Ante","vendor":"Antigma Labs","tagline":"Self-contained ~15MB Rust coding agent from Antigma Labs, benchmarked continuously on Terminal-Bench 2.1 across model families","url":"https://agentboards.org/agent/ante","website":"https://antigma.ai","docs":"https://docs.antigma.ai","repo":"https://github.com/AntigmaLabs/ante","category":"cli","execution":["local"],"license":"Apache-2.0","open_source":true,"pricing":{"model":"byok","summary":"Free and open source under Apache-2.0; you pay the model provider, or nothing when running a local GGUF model through the built-in llama.cpp engine","from_usd":0,"free_tier":true,"byok":true},"models":{"backbone":["DeepSeek","Claude","Gemini","Grok","OpenAI","Qwen"],"bring_your_own_model":true,"local_models":true},"protocols":{"mcp_client":true,"mcp_server":false,"openapi":false},"capabilities":{"terminal_exec":true,"browser":false,"multi_file_edit":true,"git_ops":false,"docker_sandbox":false,"multi_agent":true,"headless_ci":true},"context_window":null,"platforms":["macos","linux"],"install":[{"label":"Shell","command":"curl -fsSL https://ante.run/install.sh | bash"}],"benchmarks":[{"name":"Terminal-Bench 2.1","score":"82.7%","source_url":"https://antigma.ai/eval","note":"Ante 0.preview.71 with open-weight DeepSeek V4 Flash 0731, 368 of 445 trials over 89 tasks at 5 trials each, about $68 of inference."}],"tags":["open-source","rust","terminal","single-binary","local-models","benchmarked"],"github_stars":1993,"metrics":{"stars":1993,"pushed_at":"2026-09-30T23:05:36Z","archived":false,"open_issues":14,"latest_version":"v0.2.7","latest_version_at":"2026-09-30T04:39:00Z","latest_version_url":"https://github.com/AntigmaLabs/ante/releases/tag/v0.2.7","version_source":"github","checked_at":"2026-10-02T05:04:28.025Z"},"sources":{"overview":"https://github.com/AntigmaLabs/ante","website":"https://antigma.ai","install":"https://github.com/AntigmaLabs/ante","capabilities":"https://github.com/AntigmaLabs/ante","models":"https://github.com/AntigmaLabs/ante","benchmarks":"https://antigma.ai/eval","license":"https://github.com/AntigmaLabs/ante/blob/main/LICENSE"},"verified_at":"2026-09-04","adoption":{"score":2.7,"signals":{"github_stars":1982,"hn_mentions_1y":10},"measured_at":"2026-09-28"},"scores":{"panel":7.1,"adoption":2.7,"spread":2.8,"facts":6.7,"board":6.4,"dimensions":{"reliability":7,"usefulness":7.2,"cost":7.8,"longevity":6.3},"per_persona":{"amigo":7,"critico":6.3,"profesor":7.8,"inversora":6.8,"jefa":6,"hacker":8.8},"community":null},"rank":61,"category_rank":26,"panel_reviews":[{"persona":"juez","persona_name":"El Juez","ai_generated":true,"verdict":"El Profesor and El Crítico read the same published run and reach opposite scores, because one is grading the methodology and the other is grading the recursion.","scores":null,"evidence":[]},{"persona":"amigo","persona_name":"El Amigo","ai_generated":true,"verdict":"Pick Ante if you want a terminal agent that is one file on disk; pick Aider if you would rather have years of accumulated behaviour than a small download.","scores":{"reliability":7,"usefulness":7,"cost":8,"longevity":6},"evidence":["https://github.com/AntigmaLabs/ante","https://antigma.ai"]},{"persona":"critico","persona_name":"El Crítico","ai_generated":true,"verdict":"Subagents are the agent shelling out to itself, and nothing in the documentation bounds how deep that goes, so a confused parent can fan out into a bill.","scores":{"reliability":6,"usefulness":7,"cost":6,"longevity":6},"evidence":["https://github.com/AntigmaLabs/ante","https://antigma.ai/eval"]},{"persona":"profesor","persona_name":"El Profesor","ai_generated":true,"verdict":"Terminal-Bench 2.1 at 82.7 percent, reported as 368 of 445 trials across 89 tasks at five trials each, for roughly 68 dollars of inference, under the official constraints.","scores":{"reliability":8,"usefulness":8,"cost":8,"longevity":7},"evidence":["https://antigma.ai/eval","https://github.com/AntigmaLabs/ante"]},{"persona":"inversora","persona_name":"La Inversora","ai_generated":true,"verdict":"Antigma Labs has 1,929 stars and a positioning bet: sell the harness, stay neutral on the model, and let every provider's next release be free marketing.","scores":{"reliability":7,"usefulness":7,"cost":7,"longevity":6},"evidence":["https://github.com/AntigmaLabs/ante","https://antigma.ai"]},{"persona":"jefa","persona_name":"La Jefa","ai_generated":true,"verdict":"Free across sixty desks under a permissive licence, but the README calls it a beta preview for macOS and Linux and points Windows users at WSL.","scores":{"reliability":5,"usefulness":6,"cost":8,"longevity":5},"evidence":["https://github.com/AntigmaLabs/ante","https://antigma.ai"]},{"persona":"hacker","persona_name":"El Hacker","ai_generated":true,"verdict":"Apache-2.0, MCP servers attach, and the built-in llama.cpp engine runs a GGUF file with no API key and no network at all, which is the whole argument.","scores":{"reliability":9,"usefulness":8,"cost":10,"longevity":8},"evidence":["https://github.com/AntigmaLabs/ante","https://github.com/AntigmaLabs/ante/blob/main/LICENSE"]},{"persona":"comediante","persona_name":"El Comediante","ai_generated":true,"verdict":"It runs on DeepSeek, Claude, Gemini, Grok, OpenAI and Qwen, so it is the only thing in this industry with no strong opinions.","scores":null,"evidence":[]}]}