---
title: "The 10 best open-weight LLMs for coding in 2026"
description: "We rank DeepSeek V4.1 Flash, Kimi K3, GLM-5.3 and seven more open-weight models for coding agents, with licenses, sizes, vendor scores and where to run…"
author: "Ipseeta Priyadarshini"
published: "2026-10-28"
updated: "2026-10-28"
url: "https://hashnode.com/blog/best-open-weight-llms"
tags: ["AI coding","llm","open source","Deepseek"]
---

# The 10 best open-weight LLMs for coding in 2026

We ranked ten models you can download today for coding-agent work, with the license terms and vendor scores behind each pick.

**tldr:** DeepSeek V4.1 Flash is the best open-weight model for coding as of October 2026. It's MIT-licensed, has a one-million-token window, and leads Kimi K3 and GLM-5.3 by about ten points on DeepSWE in DeepSeek's own table. If you can live with custom licenses, Kimi K3 and GLM-5.3 post the highest Terminal-Bench 2.1 scores of any weights you can download, and Qwen3.8-27B fits on one machine. If you'd rather not host anything, Buildful runs four of the top five as a cloud coding agent.

This list covers the weights you can download as of 28 October 2026, and we'll update it as new models ship. The first change on the calendar is Kimi K3.1 Pro, whose weights Moonshot has promised by 4 November. Every license, size, and score below comes from the vendor's own model card. For what each host charges, see our guide to [free and cheap open-weight model APIs](https://hashnode.com/blog/free-open-weight-model-apis).

## How we ranked them

We ranked the models on three tests, in order: whether you can download the weights today, whether the license lets most teams ship, and how the vendor's own coding benchmarks look for what each model is built to do. The first test is the only thing keeping Kimi K3.1 Pro off the list.

You should read those scores with care, because every lab picks its own harness and benchmark version, and numbers from two different cards rarely line up. Moonshot reports Kimi K3 at 88.3 on Terminal-Bench 2.1 with its own Kimi Code harness, while DeepSeek's table, which runs Terminal-Bench 4.0 through Claude Code, has the same model at 12.6. So we compare models inside a single vendor's table wherever we can, and we ignore gaps of a few points between cards.

## The 10 at a glance

Each score comes from the vendor's own card and test setup, so only compare two rows where they name the same benchmark.

| # | Model | License | Total / active params | Context | Vendor coding scores |
| --- | --- | --- | --- | --- | --- |
| 1 | DeepSeek V4.1 Flash (1017) | MIT | 552B (HF counts 763B) / 8B input, 16B output | 1M | DeepSWE 77.0, Terminal-Bench 4.0 42.3 |
| 2 | Kimi K3 | Kimi K3 License (custom) | 2.8T / 104B | 1M | Terminal-Bench 2.1 88.3, DeepSWE 67.5 |
| 3 | GLM-5.3 | GLM-5.3 License (custom) | 753B (HF count) / not stated | Evaluated at 1M | Terminal-Bench 2.1 88.2, DeepSWE 66.9 |
| 4 | MiniMax M3.1 Flash | MiniMax Community License | About 428B per MiniMax (HF counts 220.5B) / 23B | 1M | DeepSWE 73.7, Terminal-Bench 4.0 38.1 |
| 5 | DeepSeek V4 Pro | MIT | 1.6T / 49B | 1M | SWE-bench Verified 80.6, SWE-bench Pro 55.4 |
| 6 | Inkling and Inkling-Small | Apache 2.0 | 975B / 41B and 276B (HF counts 266B) / 12B | Up to 1M (Inkling) | SWE-bench Verified 77.6 and 80.2 |
| 7 | Muse Spark Open | Muse Model License (custom) | 816B (HF counts 822B) / 42B | 1M | Terminal-Bench 2.1 76.4, DeepSWE 60 |
| 8 | Nemotron 3.5 Super-VL 120B-A12B | OpenMDW-1.1 | 120B / 12B | 1M | SWE-bench Verified 71.10 |
| 9 | Qwen3.8-Flash-Next | qwen-community-1.0 (custom) | 125B / 6B | 262K, extensible to 1M | SWE-bench Pro 62.5 |
| 10 | Qwen3.8-27B | Apache 2.0 | 27B dense | 262K, extensible to 1M | Terminal-Bench 2.1 73.0, SWE-bench Pro 61.7 |

## 1. DeepSeek V4.1 Flash (1017 snapshot)

**Best for:** the default model for coding agents, with no license conditions.

- **License:** MIT, OSI-approved
- **Size and context:** 552B per DeepSeek, 1M tokens
- **Coding scores:** DeepSWE 77.0 (Mini-SWE-V2 harness), Terminal-Bench 4.0 42.3 (Claude Code harness), NL2Repo 68.2
- **Run it on:** DeepSeek's API (model `deepseek-flash`), Together, Fireworks, Ollama, Cloudflare, OpenCode Go, or [Buildful](https://buildful.ai) as a coding agent

DeepSeek released V4.1 Flash on [10 September 2026](https://api-docs.deepseek.com/news/news260910) and shipped the 1017 snapshot on [17 October](https://api-docs.deepseek.com/news/news261017) with shorter outputs than the first version. The release note describes a 552B mixture-of-experts model with 8B parameters active for input and 16B for output, although Hugging Face counts 763B in the checkpoint.

DeepSeek's own [model card](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017) puts Claude Opus 5 next to it on the same three benchmarks, where Opus 5 scores 74.0, 51.8, and 75.3. Flash beats Opus 5 on DeepSWE and trails it on Terminal-Bench 4.0 and NL2Repo. Getting that close to a frontier model under an MIT license is what makes it our top pick.

## 2. Kimi K3

**Best for:** the most parameters you can download.

- **License:** Kimi K3 License, custom
- **Size and context:** 2.8T total, 104B active, 1,048,576 tokens
- **Coding scores:** Terminal-Bench 2.1 88.3, DeepSWE 67.5 (67.3 on the official leaderboard)
- **Run it on:** Moonshot's Kimi API, Together, Fireworks, DeepInfra, Ollama, or Buildful

[Kimi K3](https://huggingface.co/moonshotai/Kimi-K3) is the largest model on this list, and its Terminal-Bench 2.1 score is the highest among the models that report that benchmark. Both scores come from Moonshot's own Kimi Code harness. The weight files alone add up to about 1.56 TB, so you'll need a cluster to run it yourself.

The [license](https://huggingface.co/moonshotai/Kimi-K3/blob/main/LICENSE) opens with an MIT-style grant and then adds two gates. A model-as-a-service business past $20 million in revenue over 12 months needs a separate agreement with Moonshot, and a product past 100 million monthly users or $20 million in monthly revenue has to show "Kimi K3" in its interface. Our [launch coverage](https://hashnode.com/blog/kimi-k3-benchmark-price) has the details.

## 3. GLM-5.3

**Best for:** a coding-first model with terms that almost nobody hits.

- **License:** GLM-5.3 License, custom (MIT-style text plus the review clause)
- **Size and context:** 753B per Hugging Face, evaluated at 1M tokens
- **Coding scores:** Terminal-Bench 2.1 88.2 (Claude Code harness), DeepSWE 66.9
- **Run it on:** Z.ai's API or its GLM Coding Plan, Together, Fireworks, Ollama, Cloudflare, OpenCode Go, or Buildful

Z.ai's [card](https://huggingface.co/zai-org/GLM-5.3) calls GLM-5.3 "the most capable open-weights model for coding", and its Terminal-Bench 2.1 score sits a tenth of a point behind Kimi K3. The card doesn't give a parameter count or a context spec, though Hugging Face counts 753B and Z.ai runs its long-horizon evals with a 1M-token window.

The [license](https://huggingface.co/zai-org/GLM-5.3/blob/main/LICENSE) is an MIT-style grant with a single condition: a model-as-a-service business past $10 billion in revenue must pass Z.ai's security review before commercial use. That clause reaches almost nobody, so for nearly every team the terms work like MIT.

## 4. MiniMax M3.1 Flash

**Best for:** agents and computer use.

- **License:** MiniMax Community License, custom
- **Size and context:** about 428B total per MiniMax (Hugging Face counts 220.5B), 23B active, 1M tokens
- **Coding scores:** DeepSWE 73.7, Terminal-Bench 4.0 38.1
- **Run it on:** MiniMax's API, Together, OpenCode Go, or Buildful

MiniMax pitches M3.1 Flash at [coding, agents, and computer use](https://www.minimax.io/blog/minimax-m31f), and among the models that report DeepSWE, only DeepSeek V4.1 Flash scores higher. The [card](https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash) says the model has about 428B parameters, but the Hugging Face widget shows 220B, so check before you size your hardware.

The [license](https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash/blob/main/LICENSE) grants non-commercial use by default. Commercial use needs a "Built with MiniMax M3.1-Flash" notice and a one-time email to MiniMax. Once your product passes $20 million in yearly revenue, you need written permission first, and that applies to fine-tunes as well.

## 5. DeepSeek V4 Pro

**Best for:** the biggest model under MIT.

- **License:** MIT, OSI-approved
- **Size and context:** 1.6T total, 49B active, 1M tokens
- **Coding scores:** SWE-bench Verified 80.6, SWE-bench Pro 55.4 (Max mode)
- **Run it on:** Together, DeepInfra, Ollama, or Cloudflare

[DeepSeek V4 Pro](https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro) carries the same MIT license as Flash at three times the size. Its scores above come from Max reasoning mode, and without thinking they drop to 73.6 and 52.1. The card still calls the V4 series a preview.

It ranks below Flash because DeepSeek itself has stepped back from it. Its [10 September note](https://api-docs.deepseek.com/news/news260910) says that from 14 September, every `deepseek-v4-pro` request on DeepSeek's API goes to V4.1 Flash until V4.1 Pro launches. If you want to run Pro, you'll need someone else's GPUs.

## 6. Inkling and Inkling-Small

**Best for:** fine-tuning.

- **License:** Apache 2.0, OSI-approved
- **Size and context:** 975B / 41B (up to 1M tokens) and 276B / 12B (Hugging Face counts 266B)
- **Coding scores:** SWE-bench Verified 77.6 (Inkling) and 80.2 (Inkling-Small)
- **Run it on:** Tinker for fine-tuning (64K and 256K context options), Together, Fireworks, Modal, Databricks, Baseten, or OpenRouter's free collection

Thinking Machines built [Inkling](https://thinkingmachines.ai/news/introducing-inkling/) to be customized and released it on 15 July 2026, then shipped [Inkling-Small](https://huggingface.co/thinkingmachines/Inkling-Small) about two weeks later. On the same bash-only harness, the small model scores higher on SWE-bench Verified than the big one, which makes it the better starting point for most coding fine-tunes.

Inkling-Small is also light enough to run on one multi-GPU node. Our [Inkling fine-tuning guide](https://hashnode.com/blog/fine-tuning-llm-inkling-guide) walks through the process.

## 7. Muse Spark Open

**Best for:** coding from screenshots, recordings, or audio.

- **License:** Muse Model License, custom
- **Size and context:** 816B total (Hugging Face counts 822B), 42B active, 1M tokens
- **Coding scores:** Terminal-Bench 2.1 76.4 (Terminus 2), DeepSWE 60
- **Run it on:** Together, Fireworks, Ollama, or Buildful

Meta shipped [Muse Spark Open](https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B) on 15 October 2026, and it takes text, image, video, and audio as input. Its coding scores trail the leaders, but the license is the bigger drawback.

To download the weights, you fill in a gated form with your legal name and date of birth. Every product built on the model has to display "Built with Muse Spark Open", and a model-as-a-service business past $20 million in yearly revenue needs a separate written agreement with Meta.

## 8. NVIDIA Nemotron 3.5 Super-VL 120B-A12B

**Best for:** a permissive license at mid size.

- **License:** OpenMDW-1.1, permissive
- **Size and context:** 120B total, 12B active, 1M tokens
- **Coding scores:** SWE-bench Verified 71.10
- **Run it on:** your own GPUs from the Hugging Face checkpoints, or NVIDIA's build.nvidia.com, which hosts the Nemotron family

NVIDIA released [Nemotron 3.5 Super-VL](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16) on 15 October 2026 with vision input, and it published both BF16 and NVFP4 checkpoints. Its SWE-bench score sits well behind the top five.

What earns it a place here is the license. OpenMDW-1.1 is a permissive license written for model weights, and it has no revenue gate.

## 9. Qwen3.8-Flash-Next

**Best for:** near-frontier coding on one 128 GB machine.

- **License:** qwen-community-1.0, custom
- **Size and context:** 125B total, 6B active, 262K tokens (1M extended)
- **Coding scores:** SWE-bench Pro 62.5
- **Run it on:** your own 128 GB machine, or Alibaba's hosted Qwen3.8-Flash API, which defaults to 1M tokens

[Qwen3.8-Flash-Next](https://huggingface.co/Qwen/Qwen3.8-Flash-Next) activates only 6B parameters per token, and its SWE-bench Pro score is the highest among the models here that report one. Once you add a 51B n-gram embedding and a 4B multi-token-prediction head, the checkpoint holds 180B parameters.

The card gives no memory figure, but Ollama's tags do: the q4_K_M build is a 120 GB download and the MLX build is 105 GB. That fits a 128 GB machine with little to spare, so keep the context window modest. The license is Alibaba's own qwen-community-1.0, so read it before you ship.

## 10. Qwen3.8-27B

**Best for:** small, permissive, and local.

- **License:** Apache 2.0, OSI-approved
- **Size and context:** 27B dense, 262K tokens (1M extended)
- **Coding scores:** Terminal-Bench 2.1 73.0, SWE-bench Pro 61.7
- **Run it on:** Ollama locally, Cloudflare Workers AI, Groq (including its free plan), or Cerebras

[Qwen3.8-27B](https://huggingface.co/Qwen/Qwen3.8-27B) is a dense model under Apache 2.0. Its scores are Alibaba's, measured on a harness Alibaba chose, so don't compare them with the trillion-parameter models above. At 4 bits it's an 18 GB download in Ollama, which puts it on a single 24 GB GPU or a recent Mac. Its card also lists native image and video understanding, which helps when you want to turn a screenshot or a diagram into code.

Our companion post on [the best local LLMs for coding](https://hashnode.com/blog/best-local-llm-for-coding) covers it alongside other single-machine models.

## On the watch list: Kimi K3.1 Pro

Kimi K3.1 Pro is the next model we expect to add, as soon as Moonshot publishes the weights. Moonshot calls [K3.1 Pro](https://www.kimi.com/en/blog/kimi-k3-1-pro) its "most capable 3T-class open-weight model" and reports 65.7% on Terminal-Bench 4.0 with the Claude Code harness, against 66.4% for Claude Opus 5.5 in the same post. That would put it far ahead of the two models above that report Terminal-Bench 4.0.

The weights aren't out yet, though. Both the blog post and the [quickstart](https://platform.kimi.ai/docs/guide/kimi-k3-1-pro-quickstart.md) say they'll be released by 4 November 2026, and as of 28 October there's no K3.1 repo on Moonshot's Hugging Face page. Moonshot hasn't published a parameter count or a license either.

You can already use it through the Kimi API, OpenRouter, OpenCode Go, and Buildful.

## Open source vs open weight: what's the difference?

The difference is how much of the model you get. An open-weight model publishes its trained weights so you can download and run it. Open source, as the Open Source Initiative defines it, also asks for the code, enough detail about the training data to rebuild the model, and no limits on who may use it or for what.

By that test, a model that ships downloadable weights and nothing else counts as open-weight. Most people say "open-source LLM" for both anyway, but the license is what decides what you can build. The models on this list fall into three groups:

- **OSI-approved:** MIT (DeepSeek V4.1 Flash, DeepSeek V4 Pro) and Apache 2.0 (Inkling, Inkling-Small, Qwen3.8-27B), with no revenue gates and no branding requirements.
- **Permissive, not OSI-approved:** OpenMDW-1.1 (Nemotron 3.5 Super-VL).
- **Custom:** Kimi K3, GLM-5.3, MiniMax M3.1 Flash, Muse Spark Open, and Qwen3.8-Flash-Next, each with its own revenue thresholds, notices, or separate agreements.

The chart below turns those groups into the question most teams ask, which is whether they can ship a product on the model.

![Flowchart: what license a model is under decides whether you can ship it. MIT or Apache 2.0 means ship it with no conditions: DeepSeek V4.1 Flash, DeepSeek V4 Pro, Inkling, Inkling-Small, and Qwen3.8-27B. OpenMDW-1.1 means ship it with no revenue gate: Nemotron 3.5 Super-VL. Custom licenses mean checking the conditions first: Kimi K3, GLM-5.3, MiniMax M3.1 Flash, Muse Spark Open, and Qwen3.8-Flash-Next.](https://cdn.hashnode.com/uploads/gql/54b81b237393d8e81463c55c/e2648e39-f0d0-4175-9284-893b8709fd20.png align="center")

*Whether you can ship on each model, by license. Read the full license before you rely on it.*

## How do you code with these models without hosting them?

You can use a cloud coding agent, an API host, or a coding plan. Most teams won't run a trillion-parameter checkpoint themselves, and none of these routes asks you to.

![Flowchart: what you want back decides the route. Pull requests point to a cloud coding agent such as Buildful. An API endpoint points to an API host such as OpenRouter or Together. Usage inside your own agent points to a coding plan such as OpenCode Go or the GLM Coding Plan.](https://cdn.hashnode.com/uploads/gql/54b81b237393d8e81463c55c/1429842c-2b58-4184-bb43-13f540562b62.png align="center")

*Which route fits depends on what you want back from the model.*

**A cloud coding agent.** [Buildful](https://buildful.ai) is an AI coding agent: you describe a change, and it builds it, tests it in a real browser, and opens a pull request on your GitHub repo. It runs [open-weight models](https://buildful.ai/open-weight-models) in the cloud, so there's no GPU or API key to set up. Smart Mode picks an open-weight model for each task, and on the Pro and Team plans you can pick one yourself. The picker carries five models from this list (DeepSeek V4.1 Flash, Kimi K3, GLM-5.3, MiniMax M3.1 Flash, and Muse Spark Open) plus Kimi K3.1 Pro.

Pro costs $10 a month for 1,000 Smart Mode tasks. When you pick a model yourself, you pay from prepaid credit, and frontier models like Claude and GPT also take your own API key. The $30 [Buildful Pass](https://buildful.ai/blog/buildful-pass) gives you $80 of monthly usage on four models, Kimi K3, DeepSeek V4.1 Flash, GLM-5.3 Flash, and MiniMax M3.1 Flash, so you can pick those without topping up credit. The trade-off is that Buildful hands you pull requests rather than an endpoint, so it won't help if you need an API for your own product.

**An API host.** OpenRouter and Together serve most of the models above behind an OpenAI-compatible API, and you bring your own agent. Our [open-weight API guide](https://hashnode.com/blog/free-open-weight-model-apis) lists their prices and free tiers.

**A coding plan.** OpenCode Go and Z.ai's GLM Coding Plan sell a monthly allowance on these models for the coding agent you already use, such as Claude Code or OpenCode.

## Which one should you pick?

If you only want one, start with DeepSeek V4.1 Flash, and use the table when your constraints point elsewhere.

| If you want | Pick |
| --- | --- |
| One default model for agent work, no license conditions | DeepSeek V4.1 Flash |
| The highest Terminal-Bench 2.1 scores, and custom terms are fine | Kimi K3 or GLM-5.3 |
| An agent that drives a computer | MiniMax M3.1 Flash |
| The biggest MIT model | DeepSeek V4 Pro, from a third-party host |
| A base for your own fine-tune | Inkling-Small |
| Screenshots, video, or audio as input | Muse Spark Open |
| A permissive mid-size model | Nemotron 3.5 Super-VL |
| The strongest model on one 128 GB machine | Qwen3.8-Flash-Next |
| A laptop or single GPU, under Apache 2.0 | Qwen3.8-27B |
| A coding agent on these models, nothing to host | Buildful |

## FAQ

### What is an open-weight model?

An open-weight model is one whose trained weights are published for anyone to download, run, and fine-tune on their own hardware. The license decides what you're allowed to do with them. Open-weight doesn't mean open source, because most open-weight releases don't include the training data or the full training code.

### What's the best open-source LLM for coding?

DeepSeek V4.1 Flash is the best open-source LLM for coding as of 28 October 2026. It's MIT-licensed, has a 1M-token context, and leads Kimi K3 and GLM-5.3 by about ten points on DeepSWE in DeepSeek's own table, at 77.0 against 67.5 and 66.9. Its own Terminal-Bench 4.0 score is 42.3. If custom license terms work for you, Kimi K3 and GLM-5.3 score highest on Terminal-Bench 2.1.

### Can I use open-weight models commercially?

Usually you can, as long as you check the license first. The MIT and Apache 2.0 models (DeepSeek V4.1 Flash, DeepSeek V4 Pro, Inkling, Qwen3.8-27B) carry no revenue gates. MiniMax M3.1 Flash needs written permission once your product passes $20 million in yearly revenue, and Kimi K3 and Muse Spark Open need a separate agreement if you sell the model as a service past $20 million. MiniMax and Meta also require a "Built with" notice, and Kimi requires its name in the interface of very large products.

### Which open-weight model has the longest context?

One million tokens is the ceiling on this list. DeepSeek V4.1 Flash, DeepSeek V4 Pro, Kimi K3, MiniMax M3.1 Flash, Muse Spark Open, Nemotron 3.5 Super-VL, and Inkling all support it, while the Qwen3.8 models run 262,144 tokens natively and extend to 1M.

### What's the easiest way to use open-weight models for coding?

It depends on what you want back. If you want finished pull requests, a cloud coding agent such as [Buildful](https://buildful.ai) needs the least setup. If you want an endpoint for your own agent, an API host like OpenRouter or Together fits better, and a coding plan like OpenCode Go gives you a monthly allowance inside the agent you already use. The section on [coding without hosting](#heading-how-do-you-code-with-these-models-without-hosting-them) compares all three.