From the NVIDIA edition of October 23, 2026
Atomic Agent test: Nemotron 3.5 Bolt E4B nearly ties its Gemma 4 E4B base on agent tasks (12/18 vs 11/18) but finishes far faster; the '4GB' GGUF file is 6.18 GB
Atomic Agent (independent local-agent developer); NVIDIA; Google · Independent technical evaluation (open models) · Global
- Publisher
- Atomic Agent blog
- Original publication
- 2026-10-23
- Source retrieved
- 2026-10-24
- Event date
- 2026-10-23
What happened
Hours after release, Atomic Agent ran Bolt E4B and its base Gemma 4 E4B on six offline agent tasks, three runs each, on one RTX 5090. Bolt passed 12 of 18 runs and Gemma 11 of 18, which the authors call close to a tie; Bolt finished all runs in 113 seconds versus 662, mostly because it generated about a tenth of the tokens. VRAM at 64K context was about 4.9 GB for Bolt and 4.3 GB for Gemma; NVIDIA's Q4_K_M file was 6.18 GB. Bolt ships Gemma 4's chat template under a Nemotron name, which caused their agent to fall back to a plain mode. Licences differ: OpenMDW-1.1 for Bolt, Apache 2.0 for Gemma.
Why it matters
Useful corrective for public-sector teams evaluating small local models: the independent test shows accuracy roughly equal to the free base model, with speed gains, licence differences and integration quirks that vendor pages omit.
Evidence and measured results
Small independent test (n=3 per task, six tasks, one GPU) by a developer with its own agent product; authors note one-run differences may be noise and that they used one sampling setting for all models rather than NVIDIA's recommended temperature.
Limitations and uncertainty
Very small sample; single hardware setup; tester's own product; reasoning cap of ~500 tokens per step may affect reasoning-mode results.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-10-24; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Teams need independent evidence before choosing between similar small local models.
- Stakeholders
- IT architects
- Developers
- Open-source policy owners
- Discovery questions
- Is speed or accuracy the priority?
- Which licences are permitted?
- What hardware will run the model?
- Value hypothesis
- Interpretation: faster small models could cut latency for local assistants; accuracy gains not shown.
- Potential engagement
- Side-by-side local model evaluation on customer tasks.
- Unsupported claims
- Bolt is more accurate than Gemma 4 E4B
- Model file is 4 GB
Pre-sales engineering
Role takeaway
- Fit
- Local agent workloads on single GPUs.
- Architecture and integration
- llama.cpp or similar serving with agent frameworks that detect chat templates.
- Prerequisites
- Template-compatible framework
- Licence clearance
- Representative task set
- Constraints
- Small sample
- Single GPU test
- 64K context limit
- Security
- Keep file-write and command permissions in the agent layer.
- Proof of value
- Repeat test with more runs and recommended sampling
- Measure VRAM on target devices
Delivery
Role takeaway
- Work
- Run compatibility and accuracy testing before adoption.
- Dependencies
- Framework updates
- Licence review
- Ownership
- Development team lead.
- Skills and adoption
- Developers need evaluation methods for agent tasks.
- Governance checkpoints
- Licence approval
- Evaluation sign-off
- Acceptance criteria
- Proposed: evaluation of at least several dozen runs per model
- Proposed: template handling verified in chosen framework
- Risks
- Noise mistaken for improvement
- Licence conflicts
- Integration bugs
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Real memory use and file size exceed the '4GB' label; agent frameworks may need template handling updates.
Governance
Who approves, reviews and stays accountable for outcomes?
Licence differs from the Apache 2.0 base model, which matters for public-sector open-source policies.
Security and privacy
What data, permissions and controls need testing?
Test shows the agent framework blocking an attempt to overwrite an input file, a reminder that guardrails sit in the agent layer.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Not addressed.
Procurement
What should contracts, pricing and exit terms secure?
Licence terms (OpenMDW-1.1 versus Apache 2.0) should be reviewed before choosing between the derivative and its base model.
Operating model
Which teams own the service once it runs?
Small independent tests can be repeated in-house cheaply; agencies can run their own comparison before adoption.
Publication history
- 2026-10-23NVIDIA · Issue 486 resources
Stable resource ID: atomicagent-bolt-e4b-vs-gemma4-e4b-local-agent-test-20261023