{"id":"s5-nano","slug":"s5-nano","name":"S5-nano","orgId":"maritaca","family":"s5-nano","releaseDate":null,"availability":"ga","openWeights":true,"license":"Apache-2.0","paramsTotal":null,"paramsSource":null,"contextWindow":null,"knowledgeCutoff":null,"modelCardUrl":"https://huggingface.co/maritaca-ai/S5-nano","hfRepo":null,"modalities":["text","image"],"aliases":["S5-nano","s5-nano","s5 nano","maritaca s5-nano","maritaca-ai/S5-nano","maritaca-ai/S5-nano-FP8","maritaca-ai/S5-nano-GGUF","S5-nano-FP8","S5-nano-GGUF"],"aliasNote":null,"origin":"tracked","tracked":{"trackedSince":"2026-11-06","announced":"2026-11-02","blurb":"Maritaca AI's open-weight ~4B-parameter (≈4.2B text; HF safetensors list ~5B including vision) Portuguese/English multimodal model, continued from Qwen3.5-4B. Think (default) and Instruct modes, 262,144-token context, and an MTP layer for speculative decoding. Weights on Hugging Face as BF16, FP8 and GGUF under Apache-2.0 for self-hosting with vLLM, SGLang or llama.cpp; no Maritaca-hosted API for this model.","api":"none","openWeights":"yes","params":{"text":"~4B text (≈4.2B; HF safetensors ~5B incl. vision)","source":"Hugging Face model card"},"facts":[{"label":"Role","value":"Open-weight chat / reasoning model with Think and Instruct modes; image-text-to-text"},{"label":"Base","value":"Continued pretrain from Qwen3.5-4B; separate from the closed S5 flagship and from the Qwen3.5-4B catalog page"},{"label":"Context","value":"262,144 tokens"},{"label":"Languages","value":"Portuguese and English (HF card)"},{"label":"Checkpoints","value":"maritaca-ai/S5-nano (BF16), maritaca-ai/S5-nano-FP8, maritaca-ai/S5-nano-GGUF (BF16 / Q8_0 / Q4_K_M + mmproj)"},{"label":"Serving","value":"Self-host: vLLM, SGLang, llama.cpp. No Maritaca-hosted API stated for S5-nano"},{"label":"License","value":"Apache-2.0"},{"label":"Evals hub","value":"Observatório Maritaca (github.com/maritaca-ai/observatorio-maritaca)"}],"claims":["Lab claim: on Maritaca's own Portuguese open-model comparison (10 models; Observatório Maritaca / model-card Totals table), S5-nano Think total average is 75.25 with 23/23 coverage, above the next listed open models in that table (e.g. Qwen3.5-4B Think 68.68).","Lab claim: S5-nano Instruct total average 66.51 (23/23) on the same Maritaca comparison table.","Lab claim: Native BR Think average 83.74 and Instruct 74.65 on that table."],"sources":[{"kind":"repo","title":"maritaca-ai/S5-nano model card (Hugging Face)","url":"https://huggingface.co/maritaca-ai/S5-nano"},{"kind":"repo","title":"Observatório Maritaca (GitHub)","url":"https://github.com/maritaca-ai/observatorio-maritaca"},{"kind":"lab","title":"Maritaca AI on X: S5-nano","url":"https://x.com/MaritacaAI/status/2118781394524463564","date":"2026-11-06"},{"kind":"lab","title":"Maritaca AI on X: S5-nano weights","url":"https://x.com/MaritacaAI/status/2118768446225920029","date":"2026-11-06"}],"firstBenchmark":null},"organization":{"id":"maritaca","name":"Maritaca AI","slug":"maritaca","country":null,"url":"https://www.maritaca.ai","color":"#11110F","mark":"M"},"price":null,"answer":{"word":"Pending benchmarks.","scope":"Announced Nov 2, 2026; no independent results yet.","lead":"As of Nov 22, 2026, S5-nano is pending benchmarks on isitsota: announced Nov 2, 2026, with no independent results on the boards we track yet.","confidence":"low"},"status":"PENDING_BENCHMARKS","verdicts":[]}