Nemotron 3.5 Bolt E4B
Released on October 23, 2026
Changes from previous version
Smaller follow-on after Nemotron 3.5 Bolt: 8B total and 4B effective, text only, and a 65,536-token context. Built for local conversation and tool use in about 4GB. Bolt stays the 74B-total, 9B-active model with the 262,144-token context and the 24GB coding-agent checkpoint.
Release Summary
NVIDIA's open conversational model for local voice assistants, characters, and embedded agents. Post-trained from Gemma 4 E4B. 8 billion total parameters, 4 billion effective, text in and text out, and a 65,536-token context for both the prompt and the output. OpenMDW-1.1 weights. The Q4_K_M checkpoint is the 4GB release. On the BF16 card, MultiChallenge is 53.10, IFBench is 63.79, and BFCL-v3 is 68.51. The Q4_K_M checkpoint scores 55.60, 64.00, and 67.31 on those three. NVIDIA measures about 3.1GB to 4.3GB of memory and 52 to 133 tokens per second on RTX hardware at the short contexts in the post. A companion DFlash drafter is for speculative decoding, not a separate model. No hosted token price and no measured price per task. Smaller same-week sibling of Nemotron 3.5 Bolt.
Timeline
Nemotron 3.5 Bolt E4B released
NVIDIA publishes Bolt E4B weights on Hugging Face, including the Q4_K_M 4GB checkpoint and a DFlash drafter, for local conversational agents.