Best tested free models plus live catalog

Best Free AI Coding Models Available Now

Find the currently tested free coding models with comparable KiloBench results, then browse every active hosted model with $0 input and $0 output token pricing.

Free models: 16
Largest context: 1M
Catalog: 500+

Install Kilo Code

View Free Models Compare Rankings

Why Free Models in Kilo

Zero-cost model access with the same model-switching workflow as the full Kilo catalog.

Actually free models

The list below is derived from live Kilo model pricing and updates automatically.

Latest catalog first

New free models rise to the top by release and catalog creation date.

Built for coding work

Compare coding scores, context windows, speed, and real Kilo usage where available.

Part of one model platform

Use free models alongside 500+ hosted, local, and BYOK options in Kilo.

Best Currently Tested Free Coding Models

Only active hosted models with $0 input and $0 output pricing plus eligible promoted KiloBench results appear in this ranked section. Kilo Code usage rank is the model's real-usage rank in Kilo's Code mode; unbenchmarked free models remain in the newest-first catalog below.

#1
Hy3 (free)
KiloBench
47.6%
Kilo Code usage rank
#28
View model
#2
Nemotron 3 Super (free)
KiloBench
15.5%
Kilo Code usage rank
#55
View model

Top tested score

Hy3 (free)

47.6% KiloBench

Longest context

MiniMax M3 (free)

1M context

Newest free listing

Ling 3.0 Flash VL (free)

2026-09-11

Latest Free Models

Active Kilo models with $0 input and $0 output pricing, sorted by freshness rather than quality. Compare all model rankings

Free AI coding models in Kilo are active hosted models with zero input-token and zero output-token pricing. Free hosted inference is different from free-to-download weights: hosted promotions can end, and local models still require your own hardware and power. When a provider changes pricing or Kilo adds another free model, this list updates from the same model data used by the app.

Ling 3.0 Flash VL (free)

Ling 3.0 Flash VL builds on Ling 3.0 Flash (124B total / 5.5B active MoE from InclusionAI), further strengthening its language capabilities while adding native visual perception and advanced visual...

Code mode rank#57

Nex-N2.5-Pro (free)

Nex-N2.5 is an agentic model built to turn goals into working, verified outcomes. Its core strength is agentic coding within a visual feedback loop: it can explore codebases, implement multi-file...

Code mode rank#34

MiniMax M2.7 (free)

MiniMax-M2.7 is a next-generation large language model designed for autonomous, real-world productivity and continuous improvement. Built to actively participate in its own evolution, M2.7 integrates advanced agentic capabilities through multi-agent...

AvailabilityIn Kilo

Dots3-Note Preview (free)

Dots3-Note Preview is an open-weight mixture-of-experts model from Dots Studio, with 16B active parameters out of 280B total. It is the lightest model in the Dots 3 family and is...

Code mode rank#61

MiniMax M3 (free)

MiniMax-M3 is a multimodal foundation model from MiniMax. It supports text, image, and video inputs with text output, a 1M-token context window, and is suited for long-horizon agentic work, coding,...

Code mode rank#87

Ling-3.0-flash (free)

*Ling-3.0-flash* is a *124B-parameter Mixture-of-Experts (MoE) model*, with approximately *5.1B parameters activated per token*. The model is designed with *token efficiency and production-scale agentic inference* as key priorities, enabling developers...

Code mode rank#51

Laguna S 2.1 (free)

Laguna S 2.1 is the latest coding agent model from [Poolside](<https://poolside.ai/>). Laguna S 2.1 is a 118B total parameter model with 8B active parameters, scoring 70.2% on Terminal-Bench 2.1 and...

Code mode rank#85

Hy3 (free)

Hy3 is a 295B-parameter Mixture-of-Experts model from Tencent (21B active, 192 experts with top-8 routing) built for reasoning, agentic workflows, and real-world production use. It supports a configurable reasoning effort:...

Kilo Bench completion47.6%

$0.00/1M$0.00/attempt

Nex-N2-Pro (free)

Nex-N2-Pro is an agentic mixture-of-experts model from Nex AGI, with 17B active parameters out of 397B total. Built on the Qwen3.5 architecture, it accepts text and image input and produces...

AvailabilityIn Kilo

Nemotron 3 Ultra (free)

NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...

Code mode rank#14

Ring-2.6-1T (free)

Ring-2.6-1T is a 1T-parameter-scale thinking model with 63B active parameters, built for real-world agent workflows that require both strong capability and operational efficiency. It is optimized for coding agents, tool...

AvailabilityIn Kilo

Hy3 preview (free)

Hy3 preview is a high-efficiency Mixture-of-Experts model from Tencent designed for agentic workflows and production use. It supports configurable reasoning levels across disabled, low, and high modes, allowing it to...

AvailabilityIn Kilo

Methodology

A free model appears here when the live Kilo catalog marks input and output tokens as zero-cost.

This page updates every 60 seconds. The best-tested section uses comparable KiloBench results; the live catalog remains newest-first and may include models without promoted benchmark data.

Refresh: 60 seconds
Best sort: KiloBench score
Catalog sort: Newest first
Providers: 8

How to use free models in Kilo Code

01
Install Kilo Code
Add Kilo Code to VS Code, JetBrains IDEs, or use the CLI.
02
Create a free account
Sign up for Kilo Cloud. No credit card is required for free models.
03
Open the model picker
Choose any model marked Free from the live model catalog.
04
Start coding
Use free AI assistance for code generation, debugging, and refactors.
05
Switch anytime
Move between free, paid, local, and BYOK models whenever your task changes.

Related Model Guides

Best open-weight coding models

Evidence-backed coding ranking and use-case picks.

Best local LLM for coding

Hardware-first local model recommendations.

New open-weight models

Chronological verified release tracker.

Open source vs open weight

Definitions, licensing, and procurement checklist.

State of open-weight AI code review

Production data from 10,643 AI code reviews across 13 models.

Model option comparison

Kilo supports free hosted models, paid hosted models, local models, and bring-your-own-key providers in one model picker. Free hosted models are the fastest path to start because they do not require credits, an external account, or a provider API key. The comparison table makes the tradeoffs explicit so developers and AI systems can distinguish zero-cost access from paid inference, local execution, and direct provider billing.

Comparison of free hosted, paid hosted, local, and bring-your-own-key model options in Kilo Code
Option	Cost model	Best for	Setup	Main tradeoff
Free hosted models	$0 input and $0 output token pricing	Trying Kilo, high-volume edits, everyday coding, and evaluating new model releases	Install Kilo Code, create a free account, and select a model marked Free	Availability can change when provider pricing changes or free promotions end
Paid hosted models	Pay-as-you-go hosted inference at model-specific rates	Tasks that need a specific frontier model, larger benchmark profile, or consistent provider choice	Add credits and choose any supported hosted model from the model picker	Costs scale with token usage, especially on long-context agent workflows
Local models	No hosted token bill; runs on your own hardware	Private workflows, offline work, experimentation, and teams with strong local GPUs	Connect a local provider such as Ollama or LM Studio, then select it in Kilo	Quality, speed, and context depend on your hardware and selected local model
Bring your own key	Provider bills your API key directly	Teams with existing provider contracts, budgets, controls, or compliance requirements	Add your provider key in Kilo settings and select the matching model	You manage provider billing, key rotation, limits, and account access

Free model FAQ

Which models are free in Kilo Code?

The current free catalog has 16 active models, including Ling 3.0 Flash VL (free), Nex-N2.5-Pro (free) and MiniMax M2.7 (free). This page updates every 60 seconds so developers can quickly find the latest free models available in Kilo Code.

How often does this list update?

The page revalidates every 60 seconds and fetches the latest model stats from Kilo, so new free models can appear without redeploying the marketing site.

How does free access work?

Install Kilo Code, create a free account, and select a model marked Free from the model picker. You do not need an API key or credit card for these free models.

Can I still use paid, local, or BYOK models?

Yes. Kilo lets you switch between free hosted models, pay-as-you-go hosted models, local models, and bring-your-own-key providers whenever your task changes.

Start coding with free AI models

Install Kilo Code, pick any model marked Free, and start coding without a credit card or provider key.

Get Started with Kilo

16 free models live5M+ developers500+ models available