Models

Every model card carries its evidence limits.

Model version is currently unknown unless independently documented in the source artifacts.

Model Provider Run Completion Parse Evidence Caveat
Claude Haiku 4.5 anthropic j97fccxfgwzwrtsqcfc2xsw0ex85fpg4 100% 100% Level 3 model version unknown
Claude Opus 4.6 anthropic j978sz65j5185m8djs923vtzcd85hgk2 100% 100% Level 3 model version unknown
Claude Opus 4.7 anthropic j974xckwt2hgw6vq005x4eqhm185ffm9 100% 100% Level 3 model version unknown
Claude Sonnet 4.6 anthropic j972k3c86rfn80hmekwhvppmh185es59 100% 100% Level 3 model version unknown
Command A cohere j97emhr96a7px410xaqhvb4xfh85e6vk 100% 100% Level 3 model version unknown
Cydonia 24B V4.1 thedrummer j972f5vmnqzb67t3q6ersqjg1s85h60s 100% 100% Level 3 model version unknown
DeepSeek R1 deepseek j976q86kqd2r9w321trar7137985g7tb 100% 100% Level 3 model version unknown
DeepSeek V3.1 deepseek j97d0rz5th1hs8dt33pnk6eyvn85hr0a 100% 100% Level 3 model version unknown
DeepSeek V3.1 Terminus deepseek j975wet2gnykcvg6hbwakq6dz985hqa8 100% 100% Level 3 model version unknown
DeepSeek V3.2 deepseek j97926c87q3kq18h8k3wps5frh85eevg 100% 100% Level 3 model version unknown
DeepSeek V3.2 Exp deepseek j97bxyg56bwpk6shf2avdwr3r985g86y 100% 100% Level 3 model version unknown
DeepSeek V4 Flash deepseek j97558j95yjknfeayjqww1mjpx85e6f6 100% 100% Level 3 model version unknown
DeepSeek V4 Pro deepseek j9751m5tnrattr50rkevqc83zn85fqkb 100% 100% Level 3 model version unknown
Gemini 2.5 Flash google j97326aq5cgd6g307zqwg0vnn185kzk0 100% 100% Level 3 model version unknown
Gemini 2.5 Flash Lite google j97e691sdmdkq7aapkxmgg3cc585jxs7 100% 100% Level 3 model version unknown
Gemini 3 Flash Preview google j9703v87ezvqrq8spna4g3j9en85fyd1 100% 100% Level 3 model version unknown
Gemini 3.1 Flash Lite Preview google j9712jf329acx64c34fq8gsg5h85esx6 100% 100% Level 3 model version unknown
Gemini 3.1 Pro Preview google j974bynzat3xpze5hb0zqp130185fg7r 100% 100% Level 3 model version unknown
Gemma 4 26B A4B google j97569yb32w3gsj6hqjd6dybes85jfq7 100% 100% Level 3 model version unknown
Gemma 4 31B google j979ttg2qxqzmtn14pr5vhhr3x85j520 100% 100% Level 3 model version unknown
GLM 4.7 Flash z-ai j97adzpw9epv1m1gm2npz5p86n85kzqm 100% 100% Level 3 model version unknown
GLM 5 z-ai j97dxxae37b10j854sbjsejzwd85h6td 100% 100% Level 3 model version unknown
GLM 5.1 z-ai j975jp9v89wvjq5f3jr8nn2ewd85ekgk 100% 100% Level 3 model version unknown
Goliath 120B alpindale j972zxe0bxxkwqszyfpyckrzh185hqks 100% 100% Level 3 model version unknown
GPT OSS 120B openai j975mtsnr4s2jsxwthrxes5y5n85jbfm 100% 100% Level 3 model version unknown
GPT OSS 20B openai j97arx37czhpjcwaasmjx68p0x85j8z0 100% 100% Level 3 model version unknown
GPT-4.1 Mini openai j97cdjx35w4j8f7gw163tpvyjh85j6xt 100% 100% Level 3 model version unknown
GPT-5.2 openai j9738hkxta6qate5rm4h90nchh85e75a 100% 100% Level 3 model version unknown
GPT-5.3 Chat openai j97030dbj0smf8jgfss3cwn4p185f73k 100% 100% Level 3 model version unknown
GPT-5.3 Codex openai j97fj32eyva9ectzrn9y2fz0dh85emf3 100% 100% Level 3 model version unknown
GPT-5.4 openai j976yztmsq9f35crkvwgr6wcdn85ef9d 100% 100% Level 3 model version unknown
GPT-5.4 Mini openai j97487ck53tjz2t8zkf2fkey2n85epvs 100% 100% Level 3 model version unknown
GPT-5.4 Nano openai j97ev857pwr89j0b1c1rvbm28h85g28v 100% 100% Level 3 model version unknown
GPT-5.5 openai j97drygvx6x2mcmnygsqn2zyjs85ek57 100% 100% Level 3 model version unknown
Grok 4 Fast x-ai j976d8qr13tt14n7z0j4sf00xs85evs8 100% 100% Level 3 model version unknown
Grok 4.1 Fast x-ai j979szf5scyfhftyytsj0v9cv585eq5y 100% 100% Level 3 model version unknown
Grok 4.20 x-ai j97e8dpd84ta0958wrdey84ax185fm7p 100% 100% Level 3 model version unknown
Hermes 4 405B nousresearch j97ejpjfe0hs83r9cxy7wfatmn85hxj1 100% 100% Level 3 model version unknown
Hermes 4 70B nousresearch j973srp02sjd306vt2dyc6mfd185gvkv 100% 100% Level 3 model version unknown
Jamba Large 1.7 ai21 j975f44yspeb5es849tex4bgy185h01r 100% 100% Level 3 model version unknown
Kimi K2.5 moonshotai j970pka97k8h3fmamjx2tqspsd85eyeb 100% 100% Level 3 model version unknown
Kimi K2.6 moonshotai j972rpbb4zj92abj1wzfyszvxh85ej4k 100% 100% Level 3 model version unknown
LFM 2.5 1.2B Instruct liquid j97cyfrkh9bds0hxtk1q4y0qvh85f5xm 100% 100% Level 3 model version unknown
LFM 2.5 1.2B Thinking liquid j9779tqd3776s4xf81ns9ghzm985ejew 100% 100% Level 3 model version unknown
LFM2 24B A2B liquid j972z0ab4awj4dm83gehryxxrs85h4e3 100% 100% Level 3 model version unknown
Ling 2.6 1T inclusionai j976n5byp7xkgxwkswqhwdvkqs85exhg 100% 100% Level 3 model version unknown
Ling 2.6 Flash inclusionai j97dwh68yka3fahp3ykqz8gte585fgpr 100% 100% Level 3 model version unknown
Llama 3.3 Nemotron Super 49B V1.5 nvidia j970yx8r25xjv5cpqchbkzd2pd85hw6v 100% 100% Level 3 model version unknown
Llama 4 Maverick meta-llama j979bzkv7nt0652dfcc8a492bh85em8b 100% 100% Level 3 model version unknown
Llama 4 Scout meta-llama j97046nzsgev30rvz34023r09s85fe2r 100% 100% Level 3 model version unknown
Mercury 2 inception j974q6jv567bwj7averm73v29185gxvq 100% 100% Level 3 model version unknown
MiMo V2 Pro xiaomi j971zvzjjhvdmsja2yhwrjrdmx85f81b 100% 100% Level 3 model version unknown
MiMo V2.5 xiaomi j9798nyatbek9q6d90b3np17v985gekp 100% 100% Level 3 model version unknown
MiMo V2.5 Pro xiaomi j977zckgs0n16jkjhw18fyr15585g9yn 100% 100% Level 3 model version unknown
Ministral 3 14B 2512 mistralai j97evkvga01cbryq8k0rnqdcw585jzyf 100% 100% Level 3 model version unknown
Ministral 3 3B 2512 mistralai j974d5w0nnfmp415t12phqa3f585k252 100% 100% Level 3 model version unknown
Ministral 3 8B 2512 mistralai j9775m5heecrtm0b58p70d7f4185jya2 100% 100% Level 3 model version unknown
Mistral Large 3 2512 mistralai j97arjecv51yhx2arm6pxptpyx85fwkp 100% 100% Level 3 model version unknown
Mistral Saba mistralai j970bzqn5e2bk5bzn4pmyv8qds85g4y1 100% 100% Level 3 model version unknown
Mistral Small 4 mistralai j976b2a43zzaypq24xz28zc68d85fm62 100% 100% Level 3 model version unknown
MythoMax 13B gryphe j978d78pg1f5d9gkqkbgdns2nh85hv4p 100% 100% Level 3 model version unknown
Nemotron 3 Nano 30B A3B nvidia j97a2fb4waf743hw5f8v8dxhzs85f6rc 100% 100% Level 3 model version unknown
Nemotron 3 Super nvidia j97bpjxwb44wy95jf0ftxghj5185fcfx 100% 100% Level 3 model version unknown
Nemotron 3 Super 120B nvidia j976fzsxb60vx0vshb059v097585f2hd 100% 100% Level 3 model version unknown
Nemotron Nano 9B V2 nvidia j97d41m0hcy7zwqe15p3ha2sjs85jjr5 100% 100% Level 3 model version unknown
Nova 2 Lite amazon j97f3q6jm1m0xept1vqw1arxxd85ggb7 100% 100% Level 3 model version unknown
Nova Premier 1.0 amazon j97cfqvrcej9cr2eab6gt3n6cn85edd7 100% 100% Level 3 model version unknown
OLMo 3.1 32B Instruct allenai j9769xggfdhbzr2bz8kjgz4k9h85gqeg 100% 100% Level 3 model version unknown
OpenAI OSS 120B openai j9715xceb2fdrpdqjg380qp8nn85ftxc 100% 100% Level 3 model version unknown
OpenAI OSS 20B openai j97b9evejnbbxnvkx3eng1fg5s85f5wm 100% 100% Level 3 model version unknown
Phi 4 microsoft j973s6nx8vv8g4mvmscc0z4v7x85h828 100% 100% Level 3 model version unknown
Qwen3 Max Thinking qwen j97br8p7zfpvmanpw37gqd136x85exmc 100% 100% Level 3 model version unknown
Qwen3 Next 80B A3B Instruct qwen j971md79pynf3gevqas44y8f6h85g5ry 100% 100% Level 3 model version unknown
Qwen3.5 27B qwen j97be7gfths29n1wvwbrv0dzmh85gkee 100% 100% Level 3 model version unknown
Qwen3.5 35B A3B qwen j97adjehvznawfpb4p70ndqrxs85h8eh 100% 100% Level 3 model version unknown
Qwen3.5 397B A17B qwen j974jm92bywb9etdebymgxg8h985fnfx 100% 100% Level 3 model version unknown
Qwen3.5 Flash qwen j97evxnvs1v51qepm5yknbxv1985gcva 100% 100% Level 3 model version unknown
Qwen3.5 Plus qwen j97dh1vsp50ppzkymkx18d94zd85ffk4 100% 100% Level 3 model version unknown
Qwen3.6 Plus qwen j97da4s45y2565za5q74j76qph85f1ep 100% 100% Level 3 model version unknown
Rocinante 12B thedrummer j9719vvw14qd8pem010344rff185gjr9 100% 100% Level 3 model version unknown
Seed 2.0 Lite bytedance-seed j979jzwqhtqms8rw95c5rmmj2s85jxxj 100% 100% Level 3 model version unknown
Seed 2.0 Mini bytedance-seed j972cxx7yrkgm4kywwtcnpa0zn85gzjv 100% 100% Level 3 model version unknown
Solar Pro 3 upstage j97177wat2y8sep0zs1wh5rv2n85g9qj 100% 100% Level 3 model version unknown
Trinity Large Preview arcee-ai j974r07tcewfvap24mt4mffwj985hg7s 100% 100% Level 3 model version unknown