{"slug": "mistral-small", "name": "Mistral Small", "category": "ai-models", "category_name": "AI models", "description": "", "recommenders": 4, "commenters": 10, "warners": 1, "score": 3.715, "rising": -1.31, "first": "2025-11-14", "latest": "2026-06-16", "why": "So far it's better for equivalent Qwen 3.5 workloads, and much less expensive.", "why_is_quote": true, "has_verdict": false, "sentiment": {"recommend": 4, "neutral": 4, "compare": 1, "warn": 1}, "spark": [1, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0], "listed": true, "loved": [], "caveat": "", "short_quote": "So far it's better for equivalent Qwen 3.5 workloads, and much less expensive.", "kind": "", "icon": false, "family": "mistral", "domain": "nvidia.com", "github_repo": null, "official_url": "", "pricing": "", "aliases": ["Mistral Small 4", "Mistral Small", "Mistral Small 3.2"], "verdict": null, "praise": [{"label": "", "summary": "", "quotes": [{"id": 767349281, "item_id": 47959330, "url": "https://news.ycombinator.com/item?id=47959330", "by": "barrell", "date": "2026-04-30", "version": null, "stance": "recommend", "quote": "I've found tons of use cases for mistral small as well.", "excerpt": "Yeah I use Mistral Large for a lot of formatting work. For this one use case of mine, it outperforms frontier models by a significant margin. I've found tons of use cases for mistral small as well. I'd love to use Mistral for more tasks, but Mistral Large doesn't quite cut it for all tasks. So on the one hand, I'm excited there is another model, and presumably more performant based on the price? But the fact it's a \"", "context": "", "use_case": "", "switched_from": "", "story": "Mistral Medium 3.5"}, {"id": 758636256, "item_id": 47414766, "url": "https://news.ycombinator.com/item?id=47414766", "by": "weitendorf", "date": "2026-03-17", "version": "4", "stance": "recommend", "quote": "So far it's better for equivalent Qwen 3.5 workloads, and much less expensive.", "excerpt": "So far it's better for equivalent Qwen 3.5 workloads, and much less expensive. As you mention Qwen spends way too much time/tokens reasoning, so it ends up being more expensive than you'd think based on its model card (also IME, flaky). I actually think this model is a Big Deal because there's a whole world out there of people building on top of Qwen and other Chinese models, and now Mistral has just released one of ", "context": "", "use_case": "Building on open-weight model", "switched_from": "", "story": "Mistral Small 4"}, {"id": 737975568, "item_id": 46123473, "url": "https://news.ycombinator.com/item?id=46123473", "by": "druskacik", "date": "2025-12-02", "version": null, "stance": "recommend", "quote": "I use mistral-small with batch API and it's probably the best cost-efficient option out there.", "excerpt": "This is my experience as well. Mistral models may not be the best according to benchmarks and I don't use them for personal chats or coding, but for simple tasks with pre-defined scope (such as categorization, summarization, etc.) they are the option I choose. I use mistral-small with batch API and it's probably the best cost-efficient option out there.", "context": "", "use_case": "Simple categorization and summaries", "switched_from": "", "story": "Mistral 3 family of models released"}, {"id": 734840769, "item_id": 45927548, "url": "https://news.ycombinator.com/item?id=45927548", "by": "badsectoracula", "date": "2025-11-14", "version": null, "stance": "recommend", "quote": "it does a good enough work", "excerpt": "I've done some summarizing with my own small Tcl/Tk-based frontend that uses llama.cpp to call Mistral Small (i.e. all is done locally) and i do know that it can be off about various things. However 99% of the times i use this isn't because i need an accurate summary but because i come across some overly long article that i do not even know if i'm interested in reading, so i have Mistral Small generate a summary to g", "context": "", "use_case": "Article summaries", "switched_from": "", "story": "I think nobody wants AI in Firefox, Mozilla"}]}], "caveats": [{"label": "", "summary": "", "quotes": [{"id": 762190002, "item_id": 47636875, "url": "https://news.ycombinator.com/item?id=47636875", "by": "survirtual", "date": "2026-04-04", "version": "4", "stance": "warn", "quote": "MistralSmall4 is too buggy to use", "excerpt": "I have been running nearly the same experiments for the same reasons. It has been a lot of tweaking and patching because I am on a Strix Halo system with amd gfx1191, but the LLM component is working nicely. The evaluation I use is having it one shot a 3d scene using threejs. After everything, the output was comparable to Claude Sonnet (Opus actually does worse on this task strangely). For my local setup, I have sett", "context": "", "use_case": "local LLM", "switched_from": "", "story": "Vitalik Buterin \u2013 \"My self-sovereign / local / private / secure LLM setup\""}]}], "who": [], "tags": [], "use_cases": ["local inference", "local llm", "building on open-weight model", "agentic workflow testing", "local llm for conversation", "simple categorization and summaries"], "compared": [{"slug": "qwen", "name": "Qwen", "co": 5, "switched_from": 0, "switched_to": 0}, {"slug": "gpt-oss", "name": "gpt-oss", "co": 4, "switched_from": 0, "switched_to": 0}, {"slug": "llama-cpp", "name": "llama.cpp", "co": 2, "switched_from": 0, "switched_to": 0}, {"slug": "mistral-medium", "name": "Mistral Medium", "co": 2, "switched_from": 0, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=47404575", "title": "Mistral Small 4", "mentions": 3}, {"url": "https://news.ycombinator.com/item?id=48561269", "title": "item 48561269", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=47949642", "title": "Mistral Medium 3.5", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=47807006", "title": "Claude Opus 4.7 costs 20\u201330% more per session", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=47610534", "title": "Vitalik Buterin \u2013 \"My self-sovereign / local / private / secure LLM setup\"", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=46731068", "title": "Ask HN: What's the current best local/open speech-to-speech setup?", "mentions": 1}], "sentiment_by_month": {"2025-11": {"recommend": 1}, "2025-12": {"recommend": 1}, "2026-01": {"neutral": 1}, "2026-02": {}, "2026-03": {"compare": 1, "neutral": 1, "recommend": 1}, "2026-04": {"warn": 1, "neutral": 1, "recommend": 1}, "2026-05": {}, "2026-06": {"neutral": 1}, "2026-07": {}, "2026-08": {}, "2026-09": {}, "2026-10": {}}, "versions": [], "version_unnamed": 3, "family_ref": {"slug": "mistral", "name": "Mistral"}, "runs_on": null, "lines": [], "quotes": [{"id": 777055427, "item_id": 48565964, "url": "https://news.ycombinator.com/item?id=48565964", "by": "EagnaIonat", "date": "2026-06-16", "version": "3.2", "stance": "neutral", "quote": "mistral-small3.2", "excerpt": "> thers, like coding or more advanced agentic workflows can demand much more powerful models. You can do coding and agentic fine. For coding I use qwen3.6:35b-mlx and agentic granite4.1:3b works fine. These are the models I use. - granite4.1:3b - granite4.1:30b - gpt-oss:20b - gpt-oss:120b (less so now) - mistral-small3.2 - qwen3.6:35b-mlx There will always be use cases that don't sit on your laptop, but most of what", "context": "", "use_case": "local inference", "switched_from": "", "story": null}, {"id": 767349281, "item_id": 47959330, "url": "https://news.ycombinator.com/item?id=47959330", "by": "barrell", "date": "2026-04-30", "version": null, "stance": "recommend", "quote": "I've found tons of use cases for mistral small as well.", "excerpt": "Yeah I use Mistral Large for a lot of formatting work. For this one use case of mine, it outperforms frontier models by a significant margin. I've found tons of use cases for mistral small as well. I'd love to use Mistral for more tasks, but Mistral Large doesn't quite cut it for all tasks. So on the one hand, I'm excited there is another model, and presumably more performant based on the price? But the fact it's a \"", "context": "", "use_case": "", "switched_from": "", "story": "Mistral Medium 3.5"}, {"id": 765020723, "item_id": 47813795, "url": "https://news.ycombinator.com/item?id=47813795", "by": "KronisLV", "date": "2026-04-18", "version": "4", "stance": "neutral", "quote": "I appreciate the effort behind Qwen 3.6, Gemma 4 and Mistral Small 4", "excerpt": "The new one is quite a bit heavier! GLM 4.7 is 358B parameters: https://huggingface.co/zai-org/GLM-4.7 GLM 5.1 is 754B parameters: https://huggingface.co/zai-org/GLM-5.1 That said, 5.1 is indeed a bunch better and I could definitely see myself using it for some tasks! Sadly all of the stuff I can actually run locally is still trash (I appreciate the effort behind Qwen 3.6, Gemma 4 and Mistral Small 4 though, alongsid", "context": "", "use_case": "", "switched_from": "", "story": "Claude Opus 4.7 costs 20\u201330% more per session"}, {"id": 762190002, "item_id": 47636875, "url": "https://news.ycombinator.com/item?id=47636875", "by": "survirtual", "date": "2026-04-04", "version": "4", "stance": "warn", "quote": "MistralSmall4 is too buggy to use", "excerpt": "I have been running nearly the same experiments for the same reasons. It has been a lot of tweaking and patching because I am on a Strix Halo system with amd gfx1191, but the LLM component is working nicely. The evaluation I use is having it one shot a 3d scene using threejs. After everything, the output was comparable to Claude Sonnet (Opus actually does worse on this task strangely). For my local setup, I have sett", "context": "", "use_case": "local LLM", "switched_from": "", "story": "Vitalik Buterin \u2013 \"My self-sovereign / local / private / secure LLM setup\""}, {"id": 758636256, "item_id": 47414766, "url": "https://news.ycombinator.com/item?id=47414766", "by": "weitendorf", "date": "2026-03-17", "version": "4", "stance": "recommend", "quote": "So far it's better for equivalent Qwen 3.5 workloads, and much less expensive.", "excerpt": "So far it's better for equivalent Qwen 3.5 workloads, and much less expensive. As you mention Qwen spends way too much time/tokens reasoning, so it ends up being more expensive than you'd think based on its model card (also IME, flaky). I actually think this model is a Big Deal because there's a whole world out there of people building on top of Qwen and other Chinese models, and now Mistral has just released one of ", "context": "", "use_case": "Building on open-weight model", "switched_from": "", "story": "Mistral Small 4"}, {"id": 758543104, "item_id": 47408944, "url": "https://news.ycombinator.com/item?id=47408944", "by": "zacksiri", "date": "2026-03-16", "version": "4", "stance": "neutral", "quote": "I tested the model in an agentic workflow", "excerpt": "I tested the model in an agentic workflow. Here is the report: https://upmaru.com/llm-tests/simple-tama-agentic-workflow-q1-2026/mistral-small-4", "context": "", "use_case": "Agentic workflow testing", "switched_from": "", "story": "Mistral Small 4"}, {"id": 758536320, "item_id": 47408520, "url": "https://news.ycombinator.com/item?id=47408520", "by": "kristianp", "date": "2026-03-16", "version": "4", "stance": "compare", "quote": "Copying GPT-OSS-120b?", "excerpt": "Interesting that they target around 120 billion parameters. Just enough to fit onto a single H100 with 4 bit quant. Or 128GB APU like apple silicon, AMD AI cpus or the GB spark. Copying GPT-OSS-120b? Available to try at https://build.nvidia.com/mistralai/mistral-small-4-119b-2603", "context": "", "use_case": "", "switched_from": "", "story": "Mistral Small 4"}, {"id": 747882677, "item_id": 46742667, "url": "https://news.ycombinator.com/item?id=46742667", "by": "nsbk", "date": "2026-01-24", "version": "3.2", "stance": "neutral", "quote": "For LLM I settled with Mistral-Small-3.2-24B-Instruct-2506-Q4_K_S.gguf", "excerpt": "I'm putting together a streaming ASR + LLM + streaming TTS setup based on Nvidia speech models: nemotron ASR and magpie TTS, pipecat to glue everything together, plus an LLM of your choice. I added Spanish support using canary models, as magpie models are English-only and it still works really well. The work is based on a repo by pipecat that I forked and modified to be more comfortable to run (docker compose for the", "context": "", "use_case": "local LLM for conversation", "switched_from": "", "story": "Ask HN: What's the current best local/open speech-to-speech setup?"}, {"id": 737975568, "item_id": 46123473, "url": "https://news.ycombinator.com/item?id=46123473", "by": "druskacik", "date": "2025-12-02", "version": null, "stance": "recommend", "quote": "I use mistral-small with batch API and it's probably the best cost-efficient option out there.", "excerpt": "This is my experience as well. Mistral models may not be the best according to benchmarks and I don't use them for personal chats or coding, but for simple tasks with pre-defined scope (such as categorization, summarization, etc.) they are the option I choose. I use mistral-small with batch API and it's probably the best cost-efficient option out there.", "context": "", "use_case": "Simple categorization and summaries", "switched_from": "", "story": "Mistral 3 family of models released"}, {"id": 734840769, "item_id": 45927548, "url": "https://news.ycombinator.com/item?id=45927548", "by": "badsectoracula", "date": "2025-11-14", "version": null, "stance": "recommend", "quote": "it does a good enough work", "excerpt": "I've done some summarizing with my own small Tcl/Tk-based frontend that uses llama.cpp to call Mistral Small (i.e. all is done locally) and i do know that it can be off about various things. However 99% of the times i use this isn't because i need an accurate summary but because i come across some overly long article that i do not even know if i'm interested in reading, so i have Mistral Small generate a summary to g", "context": "", "use_case": "Article summaries", "switched_from": "", "story": "I think nobody wants AI in Firefox, Mozilla"}]}