Skip to content
OpenKey

Cheapest paid models

Paid models at $0.20 or less per 1M input tokens (provider price). 98 of 338 catalog models match as of July 2026.

98
Models
$0.010
Input from /1M
10M
Max context
May 28, 2026
Newest release

102 models on OpenKey now price under $0.20/M input tokens, and the floor keeps dropping — Ling-2.6-flash runs $0.01/M input and $0.03/M output, which is close to free for anything but the highest-volume pipelines. The newest addition, Fusion (2026-06-13), hasn't published real pricing yet, but the pack below it is dominated by Chinese labs: InclusionAI, Qwen, and ByteDance-Seed all have entries under $0.15/M input. If you're picking by pure cost, Granite 4.1 8B at $0.05/$0.10 and Ling-2.6-flash at $0.01/$0.03 are the two to benchmark first — everything else here is a tradeoff between context length, modality, and how much output you'll actually generate.

Top picks

Ling-2.6-flash

Cheapest real (non-placeholder) model in the list at $0.01/M input and $0.03/M output — on OpenKey that's $0.0103/$0.0309 after the 3% fee. 262K context, text-only. If your workload is high-volume and text-only, nothing here touches it on price.

Granite 4.1 8B

IBM's entry at $0.05/M input, $0.1/M output ($0.0515/$0.103 on OpenKey) is the cheapest model from a major lab in this list, with 131K context. Good default when you want a name-brand fallback instead of a smaller regional lab.

DeepSeek V4 Flash

$0.09/M input, $0.18/M output ($0.0927/$0.1854 on OpenKey), 1,048,576 context — the largest context window under $0.10 input in this set, plus a design_arena svg benchmark rank (23, elo 1216) showing it's actually been tested, not just cheap.

All 98 models

Full catalog →
  • Step 3.7 Flashctx 256K · inout
  • Perceptron Mk1ctx 33K · inout
  • Ring-2.6-1Tctx 262K · inout
  • Granite 4.1 8Bctx 131K · inout
  • Laguna M.1ctx 262K · inout
  • Laguna XS.2ctx 262K · inout
  • Qwen3.6 35B A3Bctx 262K · inout
  • Qwen3.6 Flashctx 1M · inout
  • DeepSeek V4 Flashctx 1.0M · inout
  • Ling-2.6-1Tctx 262K · inout
  • Hy3 previewctx 262K · inout
  • MiMo-V2.5ctx 1.0M · inout
  • Ling-2.6-flashctx 262K · inout
  • Gemma 4 26B A4B ctx 262K · inout
  • Gemma 4 31Bctx 262K · inout
  • Reka Edgectx 16K · inout
  • MiniMax M2.7ctx 205K · inout
  • GPT-5.4 Nanoctx 400K · inout
  • Mistral Small 4ctx 262K · inout
  • Nemotron 3 Superctx 1M · inout
  • Qwen3.5-9Bctx 262K · inout
  • Seed-2.0-Minictx 262K · inout
  • LFM2-24B-A2Bctx 128K · inout
  • Qwen3.5-27Bctx 262K · inout
  • Qwen3.5-35B-A3Bctx 262K · inout
  • Qwen3.5-Flashctx 1M · inout
  • MiniMax M2.5ctx 205K · inout
  • Qwen3 Coder Nextctx 262K · inout
  • Step 3.5 Flashctx 262K · inout
  • Solar Pro 3ctx 128K · inout
  • GLM 4.7 Flashctx 203K · inout
  • Seed 1.6 Flashctx 262K · inout
  • Nemotron 3 Nano 30B A3Bctx 262K · inout
  • Ministral 3 14B 2512ctx 262K · inout
  • Ministral 3 3B 2512ctx 131K · inout
  • Ministral 3 8B 2512ctx 262K · inout
  • Trinity Minictx 131K · inout
  • Olmo 3 32B Thinkctx 66K · inout
  • Voxtral Small 24B 2507ctx 32K · inout
  • gpt-oss-safeguard-20bctx 131K · inout
  • Qwen3 VL 32B Instructctx 262K · inout
  • Granite 4.0 Microctx 131K · inout
  • Qwen3 VL 8B Instructctx 256K · inout
  • Qwen3 VL 8B Thinkingctx 256K · inout
  • Qwen3 VL 30B A3B Instructctx 262K · inout
  • Qwen3 VL 30B A3B Thinkingctx 131K · inout
  • Gemini 2.5 Flash Lite Preview 09-2025ctx 1.0M · inout
  • Qwen3 VL 235B A22B Instructctx 262K · inout
  • Qwen3 Coder Flashctx 1M · inout
  • Qwen3 Next 80B A3B Instructctx 262K · inout
  • Qwen3 Next 80B A3B Thinkingctx 262K · inout
  • Qwen3 30B A3B Thinking 2507ctx 131K · inout
  • Hermes 4 70Bctx 131K · inout
  • GPT-5 Nanoctx 400K · inout
  • gpt-oss-120bctx 131K · inout
  • gpt-oss-20bctx 131K · inout
  • Qwen3 Coder 30B A3B Instructctx 160K · inout
  • Qwen3 30B A3B Instruct 2507ctx 131K · inout
  • Qwen3 235B A22B Thinking 2507ctx 262K · inout
  • GLM 4.5 Airctx 131K · inout
  • UI-TARS 7B ctx 128K · inout
  • Gemini 2.5 Flash Litectx 1.0M · inout
  • Qwen3 235B A22B Instruct 2507ctx 262K · inout
  • Hunyuan A13B Instructctx 131K · inout
  • Mistral Small 3.2 24Bctx 128K · inout
  • Gemma 3n 4Bctx 33K · inout
  • Llama Guard 4 12Bctx 164K · inout
  • Qwen3 14Bctx 132K · inout
  • Qwen3 30B A3Bctx 131K · inout
  • Qwen3 32Bctx 131K · inout
  • Qwen3 8Bctx 131K · inout
  • GPT-4.1 Nanoctx 1.0M · inout
  • Llama 4 Maverickctx 1.0M · inout
  • Llama 4 Scoutctx 10M · inout
  • Gemma 3 12Bctx 131K · inout
  • Gemma 3 4Bctx 131K · inout
  • Gemma 3 27Bctx 131K · inout
  • GPT-4o-mini Search Previewctx 128K · inout
  • Reka Flash 3ctx 66K · inout
  • Sabactx 33K · inout
  • Mistral Small 3ctx 33K · inout
  • MiniMax-01ctx 1.0M · inout
  • Phi 4ctx 16K · inout
  • Command R7B (12-2024)ctx 128K · inout
  • Llama 3.3 70B Instructctx 131K · inout
  • Nova Lite 1.0ctx 300K · inout
  • Nova Micro 1.0ctx 128K · inout
  • Qwen2.5 7B Instructctx 131K · inout
  • Llama 3.2 1B Instructctx 131K · inout
  • Llama 3.2 3B Instructctx 131K · inout
  • Command R (08-2024)ctx 128K · inout
  • Llama 3 8B Lunarisctx 8K · inout
  • Llama 3.1 8B Instructctx 131K · inout
  • Mistral Nemoctx 131K · inout
  • GPT-4o-minictx 128K · inout
  • GPT-4o-mini (2024-07-18)ctx 128K · inout
  • Llama 3 8B Instructctx 8K · inout
  • MythoMax 13Bctx 4K · inout

Prices per 1M tokens, flat 3% fee included. Hover or tap a price for the math.

How to choose

First tradeoff: input vs. output price. Ministral 3 3B and Reka Edge price input and output identically ($0.1/$0.1), which is unusual — most models here charge 3-8x more for output (Qwen3.5-27B: $0.195 in / $1.56 out). If your workload is output-heavy (long completions, code generation), a model with a narrow in/out spread beats a low headline input price. Second: context length varies from 16,384 (Reka Edge) to 1,048,576 (DeepSeek V4 Flash, MiMo-V2.5) at similar price points — don't pay for 1M context you won't use. Third: modality. Text-only models (Ling-2.6-flash, Nemotron 3 Nano) are consistently cheaper than the text+image+video models (Qwen3.6 Flash, MiMo-V2.5) doing the same job, so drop multimodal support if you don't need it.

Questions

What's the actual cheapest model on OpenKey right now?
Ling-2.6-flash from InclusionAI, at $0.01/M input and $0.03/M output provider-side. On OpenKey that's $0.0103/$0.0309 per 1M tokens after the flat 3% fee. It's text-only with a 262,144-token context window, released 2026-04-21.
Are there free models on OpenKey?
Yes — 25 of the 329 total models on OpenKey are free. This page covers the 102 paid models with the lowest list prices, not the free tier. If your workload can tolerate free-tier limits, check the free models collection first.
Why do some models show negative or missing prices?
A few entries (Fusion, Pareto Code Router, Body Builder) list placeholder pricing in the raw data and haven't published real per-token rates yet. Don't budget against them — treat them as unpriced until the provider confirms rates.