MODEL · INTERNSCIENCE (SHANGHAI AI LAB) · 35B TOTAL / 3B ACTIVE (MOE FLAGSHIP) · 4.5B DENSE (AGENTS-A1-4B)
Agents-A1 (35B-A3B + 4B)
An agent-specialist family post-trained on Qwen3.5 bases for long-horizon search, engineering, scientific research, instruction following, and tool calling. The pitch is agent-horizon scaling rather than parameter scaling: InternScience reports the 35B-A3B matching much larger models on agentic benchmarks, and the 4B — released July 14 — posting large gains over its own Qwen3.5-4B base (BrowseComp 66.8 vs 47.2, MatTools 49.3 vs 10.9). Unusually for a lab release, the evaluation framework is open-sourced so the numbers are reproducible.
License: Apache 2.0 · Context: 262K · Released: June 26, 2026 (35B-A3B); July 14, 2026 (4B)
The decision in five lines
- The call
- Consider — runnable locally, family reference
- Best for
- Local evaluation and family reference
- Runs on
- 23 hardware picks fit (cheapest: Intel Arc B580 12 GB · $249)
- Watch out
- General chat or coding — this is an agent specialist, not a daily driver (the system prompt itself tells the model to answer directly without tools for ordinary questions).
- Evidence
- Estimated
- 35B total
- PARAMETERS
- AGENTIC SPECIALIST
- TYPE
- 262K
- CONTEXT
- ~20 GB (35B-A3B) / ~3 GB (4B) — the 4B runs on almost anything
- VRAM AT Q4
Where we recommend this
This model isn’t currently in an active planner slot. See the runner notes below if you’re running it anyway.
The call
An agent-specialist family post-trained on Qwen3.5 bases for long-horizon search, engineering, scientific research, instruction following, and tool calling. The pitch is agent-horizon scaling rather than parameter scaling: InternScience reports the 35B-A3B matching much larger models on agentic benchmarks, and the 4B — released July 14 — posting large gains over its own Qwen3.5-4B base (BrowseComp 66.8 vs 47.2, MatTools 49.3 vs 10.9). Unusually for a lab release, the evaluation framework is open-sourced so the numbers are reproducible.
When not to use: General chat or coding — this is an agent specialist, not a daily driver (the system prompt itself tells the model to answer directly without tools for ordinary questions). All headline numbers are vendor-run, and the "trillion-parameter performance" framing is marketing: these are post-trains of Qwen3.5, not from-scratch models. Verify on your own agent loop before displacing a proven pick.
Runner notes
Official GGUFs (Q4_K_M / Q8_0 / F16) and FP8 from InternScience itself, plus mlx-community quants for Apple Silicon — so llama.cpp, Ollama, and LM Studio all work day one. vLLM/SGLang serve both sizes at 262K context on a single GPU; pass `--tool-call-parser qwen3_coder` for tool use, and `--language-model-only` on vLLM to skip the vision encoder and free KV-cache memory.
Hardware that fits
Every hardware pick whose memory fits this model at the quant we recommend. Sorted cheapest-first — the top row is your best-value fit. Click through for the full buyer’s guide.
- Intel Arc B580 12 GBPerfect · 2.7× 12 GB · $249–$299
- NVIDIA RTX 3060 12 GBPerfect · 2.7× 12 GB · $280–$400
- Minisforum UM890 ProPerfect · 5.3× 32 GB DDR5 (shared) · $463–$580 all-in
- RTX 5060 Ti 16 GBPerfect · 3.6× 16 GB · $560–$610
- AMD Radeon RX 9070 XTPerfect · 3.6× 16 GB · $649–$779
- Mac Mini M4 16 GBPerfect · 2.4× 16 GB unified · $799 (new floor) / $499–$599 (eBay/residuals)
- AMD Radeon RX 7900 XTXPerfect · 5.3× 24 GB · $810 used / ~$1,340 new
- NVIDIA RTX 3090 (used, single)Perfect · 5.3× 24 GB · $950–$1,200
- NVIDIA RTX 5070 TiPerfect · 3.6× 16 GB · $980–$1,300
- NVIDIA RTX 5080Perfect · 3.6× 16 GB · $1,250–$1,400
- MacBook Air M5 24 GBPerfect · 3.6× 24 GB unified · $1,499–$1,899
- Mac Mini M4 Pro 24 GBPerfect · 3.6× 24 GB unified · $1,599
- Dual RTX 3090 (used)Perfect · 10.7× 48 GB · $1,800–$2,500 all-in
- NVIDIA RTX 4090Perfect · 5.3× 24 GB · $2,200–$2,800
- Framework Desktop (Ryzen AI Max+ 395)Perfect · 19.1× 128 GB unified · $2,459–$2,851 (128 GB config)
- M5 Pro MacBook Pro 48 GBPerfect · 7.1× 48 GB unified · $2,999–$3,599
- NVIDIA RTX 5090Perfect · 7.1× 32 GB · $3,500–$4,300
- NVIDIA RTX A6000 (48 GB, used)Perfect · 10.7× 48 GB ECC · $3,500–$4,500
- Mac Studio M4 Max 64 GBPerfect · 9.5× 64 GB unified · $3,799
- NVIDIA DGX SparkPerfect · 19.1× 128 GB unified · $4,699
- M5 Max MacBook Pro 64 GBPerfect · 9.5× 64 GB unified · ~$5,199 (est.; June 25 2026 increase)
- Mac Studio M3 Ultra 96 GBPerfect · 14.3× 96 GB unified · $5,299
- Dual RTX 5090Perfect · 14.2× 64 GB (2×32) · $8,500–$10,500
Next step
Find-by-model — see what hardware runs this→