NVIDIA's chart: Nemotron Bolt E4B tops small models at ~320ms first-token latency
NVIDIAAI · x · 2026-10-23
In a follow-up reply, NVIDIA shared a benchmark chart for Nemotron 3.5 Bolt E4B 4GB: it scores 70.2 at 320ms time-to-first-token, landing in the "smart and fast" zone — beating bigger-memory peers like Gemma 4 E4B (63.2) and Qwen 3.5 9B (64.1), and trailing only Gemma 4 26B-A4B (72.8) and Qwen 3.8 27B (67.0). Circle size shows VRAM usage; Bolt needs only 4GB.
Related event: NVIDIA Releases Nemotron 3.5 Bolt E4B for On-Device Agents(2 posts)→
More from Models
- Community note debunks claim that distant tokens never share MoE experts — NathanpmYoung · 2026-10-24
- FreshStack retrieval leaderboard revamped with Cohere Embed v5 and Perplexity models — beirmug · 2026-10-24
- Kimi K3.1 Pro flagship model spotted rolling out on Kimi web and mobile — testingcatalog · 2026-10-24
- NVIDIA: Bolt E4B tops MultiChallenge at 56.03% with weights, data, recipes open — NVIDIAAI · 2026-10-24
- Qwen3.8-27B at 1.8bpw collapses to 13% on MiMo RL bench, EXL3 tests show — casper_hansen_ · 2026-10-24
- Ablating Jev-style decision models: LoRA-only hits 84.9, beating the full recipe — vanstriendaniel · 2026-10-24