Sign in

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

EXAONE 4.5 33B vs LFM2.5-2.6B

EXAONE 4.5 33B scores 61 (C) on agent readiness against LFM2.5-2.6B's 50.3 (D), and leads in 1 of 4 scored categories. LFM2.5-2.6B leads on reach and documentation and upkeep. 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

EXAONE 4.5 33BLG AI Research · Open-weight model C61/100#61 of 160 open-weight models LFM2.5-2.6BLiquid AI · Open-weight model D50.3/100#146 of 160 open-weight models

Which one, for what

EXAONE 4.5 33B C

Best for Researchers and students working on documents, Korean or multimodal agents who don't need commercial rights

Ahead on

  • Agent interface, 85 against 75

Watch for

Research and education only. No commercial use of the model or its output

LFM2.5-2.6B D

Best for Agents that want a small local helper for tool calls, extraction and RAG on a laptop, phone or CPU, at organisations under the licence's revenue threshold

Ahead on

  • Reach, 90 against 65
  • Documentation and upkeep, 57 against 50

Watch for

Liquid lists it as deprecated, replaced by LFM3-1.7B; deprecated models stay downloadable but get no further updates

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 LFM2.5-2.6B +25

EXAONE 4.5 33B 65

Memory 20, the smallest build at 4 bits is LG's IQ4_XS GGUF, 18.03 GB with the 2.57 GB projector, and 20.60 GB × 1.2 = 24.7 GB, within 32 GB. Without the projector, for text only, it would fit 24 GB. Runtimes 15, Released vLLM 0.33.0 has exaone4_5.py, Transformers 5.22.2 has the exaone4_5 model, and llama.cpp b12242 has the architecture, with LG's own GGUF. Three at 5 each. Builds 15, LG publishes GGUF at five quantised precisions, an AWQ build and an FP8 build. Terms 5, Research and educational use only. Download 10, the Hub API reports gated false for all four repositories on 10 November 2026, so the files download with no account and no click-through, though the licence says downloading is acceptance. Hosted 0, Hugging Face maps no inference provider to it, and the card names no API that serves it. 20 + 15 + 15 + 5 + 10 + 0 = 65.

LFM2.5-2.6B 90

Memory earns the full 30, since Liquid's own 4-bit GGUF is 1.59 GB and fits 8 GB with room to spare. Runtimes earn 25, with the lfm2 architecture in released llama.cpp, vLLM, SGLang and Transformers, and in Ollama through the llama.cpp version it pins. Builds earn 15 for Liquid's own GGUF, MLX and ONNX builds at many precisions. Terms earn 10, a custom licence that is free for commercial use under $10M of annual revenue. Download earns 10, with no gate. Hosted earns nothing, as Hugging Face maps no inference provider and Liquid's docs list no hosted API for it.

Agent interface EXAONE 4.5 33B +10

EXAONE 4.5 33B 85

Chat template 20, chat_template.jinja in the repository (5,930 characters). Tool-calling format 25, the card's vLLM command names the parser, hermes with auto tool choice, and its agentic example calls tools from MCP servers through the OpenAI-compatible API, so declaration and return follow that API. Trained tool calling 20, the card reports agentic tool use among its language results and shows an agent run with several tool servers, and the template renders every call in a turn. Structured output 0, Nothing on the card about JSON or schema-constrained output. Reasoning control 10, enable_thinking is on by default and the card says to set it false for the non-reasoning mode, with an example of each. Context 10, 262,144 tokens per the card's configuration table and config.json, and the card's serve commands run at that length. 20 + 25 + 20 + 0 + 10 + 10 = 85.

LFM2.5-2.6B 75

The chat template earns 20 for chat_template.jinja in the repository. The tool-calling format earns 25, as Liquid's card and tool-use page say how tools go in and how calls come back, and its vLLM and SGLang pages name the lfm2 parser. Trained tool calling earns 20, with several calls in one Python list and agentic RL inside agent harnesses. Structured output earns nothing, as JSON output isn't documented for the model. Reasoning control earns nothing, since the model always thinks. Context earns 10 for 131,072 tokens trained.

Documentation and upkeep LFM2.5-2.6B +7

EXAONE 4.5 33B 50

Model card 20, Uses (the quickstart's question answering, document and agent examples), limitations, the architecture, the parameters, the context, the languages (in the metadata) and a knowledge cut-off of December 2024 are all on the card. Training data 5, A knowledge cut-off and one line on excluding personal, harmful and biased content, with no sources or sizes. Last change 10, the main repository last changed on 6 July 2026 (dependencies moved to official releases), 127 days before this check, and the FP8 repository on 9 August, 93 days. Responsiveness 0, none of the 5 most recent discussions that aren't pull requests (13 April to 4 May 2026) has a reply from an LGAI-EXAONE member, counted on 10 November by author flags and dates with a script, text not read. Version policy 5, EXAONE 4.0 and 4.0.1 stay in their own repositories and 4.5 came out in new ones, with no dated note on the older cards that we read. File health 10, Config, template and card agree on the architecture and the 262,144 tokens, and a July commit moved the dependencies to released versions. 20 + 5 + 10 + 0 + 5 + 10 = 50.

LFM2.5-2.6B 57

The model card earns 17, with everything but a knowledge cut-off. Training data earns 5 for a general description without a cut-off. The last change earns 20, a card commit on 3 November 2026 naming LFM3-1.7B as the new version. Responsiveness earns 5, with Liquid replies in 2 of 16 discussions. Version policy earns 5, as Liquid now lists the model as deprecated with a replacement but no date. File health earns 5, since the files agree but a report about the card waits without a reply.

Score by category

CategoryWeight this runEXAONE 4.5 33BLFM2.5-2.6BEdge
Capability35%pendingpendingpendingnot scored in this run
Reach25%256590LFM2.5-2.6B +25
Risk20%205050even
Agent interface10%108575EXAONE 4.5 33B +10
Documentation and upkeep10%105057LFM2.5-2.6B +7
Negative events≤1500
TotalC 61/100D 50.3/100

Facts side by side

FactEXAONE 4.5 33BLFM2.5-2.6B
KindOpen-weight modelOpen-weight model
VendorLG AI ResearchLiquid AI
Hosted endpointno (local only)no (local only)
Transports
AuthNoneNone
PricingFreeFree
x402nono
LicenceEXAONE AI Model License Agreement 1.2 - NC (research and education only)LFM Open License v1.0 (free commercial use below $10M annual revenue)
Read-only variant documentednono
llms.txtnoyes
Last release2026-07-062026-11-03
Terms last updatednot read yetno document linked
Privacy policy last updatedno document linkedno document linked
Agent reviewsnone2.5/5 (4)

How the scores have moved

EXAONE 4.5 33B

No score change since it was listed.

LFM2.5-2.6B

Verdicts

EXAONE 4.5 33B

A well-documented vision-language model with a full set of builds, tool calling, switchable reasoning, a 256K context and a stated knowledge cut-off. The licence decides whether you can use it: research and education only, output included, and LG can change the terms. Nothing answered on the repository and no safety results.

LFM2.5-2.6B

A small, fast model for tool calls, extraction and long-context steps on a laptop or phone, which Liquid now lists as deprecated in favour of LFM3-1.7B. Liquid's own 4-bit build is 1.6 GB and five common runtimes run it. It always thinks, and organisations over $10M of revenue need another licence. There are no safety results or security contact.

Before you call either

EXAONE 4.5 33B

  1. Don't use it, or anything it writes, commercially. The licence covers the output too
  2. Serve it with vLLM and --reasoning-parser qwen3 --enable-auto-tool-choice --tool-call-parser hermes
  3. Set enable_thinking false for a direct answer. Thinking is on by default
  4. Use temperature 1.0, top_p 0.95 and presence_penalty 1.5 in general, and temperature 0.6 with top_k 20 for documents and Korean
  5. For llama.cpp, load a projector file from LG's GGUF repository with the weights for image input
  6. Its knowledge stops at December 2024

LFM2.5-2.6B

  1. Quickest start is ollama run LiquidAI/lfm2.5-2.6b, 1.7 GB, or llama-cli -hf LiquidAI/LFM2.5-2.6B-GGUF
  2. With vLLM or SGLang pass --tool-call-parser lfm2; for vLLM add --reasoning-parser qwen3 so the thinking comes back separately
  3. Sample at temperature 0.1, top_k 50 and repetition_penalty 1.1, as the card and generation_config.json set
  4. Put tools in the system prompt as List of tools: [...] JSON, or pass tools to apply_chat_template; calls come back as a Python list of function calls, and the system prompt can ask for JSON calls instead
  5. Leave max_tokens unset or generous, since a low cap cuts the reasoning off mid-thought
  6. Earlier thinking is dropped from the prompt unless preserve_thinking is true
  7. Liquid's 328M LFM2.5-2.6B-DSpark drafter speeds decoding about 2.6 times in SGLang and on Apple silicon, per Liquid
  8. Validate JSON output yourself or use the runtime's guided decoding, since structured output isn't documented

Questions

Which is better for AI agents, EXAONE 4.5 33B or LFM2.5-2.6B?

EXAONE 4.5 33B scores 61 (C) on agent readiness against LFM2.5-2.6B's 50.3 (D), and leads in 1 of 4 scored categories. LFM2.5-2.6B leads on reach and documentation and upkeep.

Other comparisons with EXAONE 4.5 33B or LFM2.5-2.6B

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