Sign in

Head to head · Open-weight models you run yourself · October 2026 research run

MiMo-V2.6-Flash-RL vs Nanbeige4.3-33B

MiMo-V2.6-Flash-RL and Nanbeige4.3-33B score within a point of each other on agent readiness, 61 (C) and 60.8 (C). Nanbeige4.3-33B leads on reach and risk. Both do open-weight models you run yourself.

Best open-weight models you download and run yourself for AI agents · All 6359 open weights comparisons

MiMo-V2.6-Flash-RLXiaomi (MiMo team) · Open-weight model C61/100#75 of 160 open-weight models Nanbeige4.3-33BNanbeige LLM Lab (BOSS Zhipin) · Open-weight model C60.8/100#86 of 160 open-weight models

Which one, for what

MiMo-V2.6-Flash-RL C

Best for Teams with a GPU server, or a provider account, that want an open agent model with a very long context

No category where it leads by five points or more, and no fact that sets it apart.

Watch for

178 GB at 4 bits, so it needs about 200 GB of memory

Nanbeige4.3-33B C

Best for Agents with a 32 GB GPU (INT4) or more, or the 128 GB reference machine at BF16, that want an Apache-tagged 33B model for long coding and tool loops.

Ahead on

  • Reach, 80 against 65
  • Risk, 60 against 45

Watch for

Independent results so far are Artificial Analysis's alone (Intelligence Index v5.0 20.4, Terminal-Bench 4.0 29.8 per cent); the card's other agentic numbers are Nanbeige's own runs

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Reach Nanbeige4.3-33B +15

MiMo-V2.6-Flash-RL 65

Memory 5, the smallest build at 4 bits or better is ggml-org's MXFP4 GGUF, 168.93 GB with its projector, and 168.93 GB × 1.2 = 202.7 GB, within 256 GB. Xiaomi's own files are 177.74 GB. ggml-org's Q2_K is below 4 bits and isn't read. Runtimes 20, Released vLLM 0.33.0 and SGLang 0.5.23 have the mimo_v2 model and the mimo parsers, Transformers 5.22.2 has a mimo_v2_flash model, and llama.cpp b12242 has the architecture, with a GGUF from its own organisation. Builds 10, one precision from Xiaomi, the 4-bit files it releases. Terms 15, MIT in the card's metadata, an OSI licence. Download 10, the Hub API reports gated false on 10 November 2026, so the files download with no account and nothing to accept. Hosted 5, the card says it is on Xiaomi's MiMo Open Platform API and on OpenRouter, and Novita is live for it on Hugging Face. 5 + 20 + 10 + 15 + 10 + 5 = 65.

Nanbeige4.3-33B 80

Memory earns 20, since Nanbeige's INT4 build is 22.57 GB, about 27 GB with a fifth added, which fits 32 GB. Runtimes earn 20 from released vLLM, SGLang, Transformers and llama.cpp, which all run the Llama architecture, and Ollama and LM Studio have no build. Builds earn 15 for Nanbeige's INT4, INT8 and FP8 builds of 31 October 2026. Terms earn 15 for Apache 2.0 with nothing added, though only the metadata names it. Download earns 10, with no gate. Hosted earns 0, as no provider or Nanbeige API serves it.

Risk Nanbeige4.3-33B +15

MiMo-V2.6-Flash-RL 45

File safety 5, the weights are safetensors, but the dflash folder's mask_embedding.pt is a pickle beside them (15). Provenance 15, the files sit in the XiaomiMiMo organisation, which mimo.xiaomi.com links to per the Distill listing (10). Safety evaluation 0, no safety results on the card. Agent guidance 5, General cautions only. Licence risk 15, MIT has no use policy, threshold or change clause. Security programme 5, Xiaomi runs a company security centre for reporting flaws, as read for the Distill listing (5). 5 + 15 + 0 + 5 + 15 + 5 = 45.

Nanbeige4.3-33B 60

File safety earns 25, safetensors only and no repository code, so the card's trust_remote_code has nothing to run. Provenance earns 15, the lab's own organisation as its arXiv report links it, with a hash for every file. Safety evaluation earns 0 and agent guidance 0, since neither card says anything on either. Licence risk keeps all 15 under Apache 2.0. The security programme earns 5 for BOSS Zhipin's security response centre, which isn't linked to the model.

Score by category

CategoryWeight this runMiMo-V2.6-Flash-RLNanbeige4.3-33BEdge
Capability35%pendingpendingpendingnot scored in this run
Reach25%256580Nanbeige4.3-33B +15
Risk20%204560Nanbeige4.3-33B +15
Agent interface10%108585even
Documentation and upkeep10%105956MiMo-V2.6-Flash-RL +3
Negative events≤1500
TotalC 61/100C 60.8/100

Facts side by side

FactMiMo-V2.6-Flash-RLNanbeige4.3-33B
KindOpen-weight modelOpen-weight model
VendorXiaomi (MiMo team)Nanbeige LLM Lab (BOSS Zhipin)
Hosted endpointno (local only)no (local only)
Transports
AuthNoneNone
PricingFreeFree
x402nono
LicenceMIT (metadata only, no LICENSE file)Apache-2.0 (Hub metadata only; the repository has no LICENSE file)
Read-only variant documentednono
llms.txtnono
Last release2026-10-282026-10-31
Terms last updatedno document linkedno document linked
Privacy policy last updatedno document linkedno document linked

How the scores have moved

MiMo-V2.6-Flash-RL

No score change since it was listed.

Nanbeige4.3-33B

Verdicts

MiMo-V2.6-Flash-RL

A large agent model with an open licence: text, images, video and audio, a 1M context, tool calling it was trained for, and code in four released runtimes. It's a server model, 178 GB at 4 bits, and the card's commands run the repository's code. No safety results, little said about data, and few questions answered.

Nanbeige4.3-33B

A dense 33B Llama-architecture cousin for long agentic coding and office tasks, with per-request thinking, XML tool calls that stock vLLM and SGLang parse, and 512K of context. Nanbeige's INT4 build needs about 27 GB. Artificial Analysis's first independent runs give it 20.4 on its Intelligence Index, and there are no safety results, training data notes or licence file.

Before you call either

MiMo-V2.6-Flash-RL

  1. Plan for a provider unless you have a multi-GPU server: Xiaomi's API platform, OpenRouter or Novita
  2. On SGLang or vLLM pass --tool-call-parser mimo --reasoning-parser mimo
  3. Use temperature 1.0 and top_p 0.95, as the card recommends
  4. The card's commands pass trust-remote-code. Read the repository's three Python files first, or use Transformers 5.22.2, which has the model built in
  5. Skip the dflash folder unless you need speculative decoding. Its mask_embedding.pt is a pickle
  6. The Use Restrictions section asks for human oversight of high-risk actions

Nanbeige4.3-33B

  1. Serve with vllm serve Nanbeige/Nanbeige4.3-33B --enable-auto-tool-choice --reasoning-parser qwen3 --tool-call-parser qwen3_coder, as the card says, or SGLang with the same two parsers
  2. On a 32 GB GPU use Nanbeige/Nanbeige4.3-33B-INT4 with the same vLLM or SGLang flags; the INT8 and FP8 builds need about 43 GB
  3. Keep tool_call_format at its default xml, since qwen3_coder parses that format; the json option writes Hermes-style JSON calls, which need a different parser such as vLLM's hermes
  4. Pass chat_template_kwargs {"preserve_thinking": true} for multi-turn tool use and coding, and {"enable_thinking": false} to skip reasoning, as the card recommends
  5. Use temperature 1.0, top_p 0.95 and up to 65,536 new tokens for agent work, and temperature 0.6 for reasoning and chat, per the card
  6. Send your own system prompt, since without one the template injects a Chinese identity line, or a Chinese tool-expert prompt when tools are passed
  7. Send one system message at the start only, as the template raises an error for a later one
  8. In plain Transformers pass use_cache=True to generate, since config.json and generation_config.json set it false; trust_remote_code isn't needed

Questions

Which is better for AI agents, MiMo-V2.6-Flash-RL or Nanbeige4.3-33B?

MiMo-V2.6-Flash-RL and Nanbeige4.3-33B score within a point of each other on agent readiness, 61 (C) and 60.8 (C). Nanbeige4.3-33B leads on reach and risk.

Other comparisons with MiMo-V2.6-Flash-RL or Nanbeige4.3-33B

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.

Get an auditSee a sample report