Price per 1M logo

Cheapest tool-calling models

Models that support function calling, ranked by blended cost — the backbone of agent loops and structured extraction.

Current leader: Gemini Exp 1114 (Google Gemini). Refreshed 2026-09-13 13:40 UTC.

#ModelProviderInput $/1MOutput $/1MCached $/1MContext
1Gemini Exp 1114Google Geminifreefree1.0M
2Gemini Exp 1206Google Geminifreefree2.1M
3Gemma 3 27b ItGoogle Geminifreefree131K
4Gemma 4 26b A4b ItGoogle Geminifreefree262K
5Gemma 4 31b ItGoogle Geminifreefree262K
6Learnlm 1.5 Pro ExperimentalGoogle Geminifreefree33K
7Labs Leanstral 1 5Mistral AIfreefree262K
8Labs Leanstral 1 5 1Mistral AIfreefree262K
9Anthropic.claude Mythos PreviewAmazon Bedrockfreefree1M
10Deepseek Coder V2 BaseOllamafreefree8K
11Deepseek Coder V2 InstructOllamafreefree33K
12Deepseek Coder V2 Lite BaseOllamafreefree8K
13Deepseek Coder V2 Lite InstructOllamafreefree33K
14Deepseek V3.1:671b CloudOllamafreefree164K
15GPT Oss:120b CloudOllamafreefree131K
16GPT Oss:20b CloudOllamafreefree131K
17Internlm2 5 20b ChatOllamafreefree33K
18Llama3.1Ollamafreefree8K
19MistralOllamafreefree8K
20Mistral 7B Instruct V0.1Ollamafreefree8K
21Mistral 7B Instruct V0.2Ollamafreefree33K
22Mistral Large Instruct 2407Ollamafreefree66K
23Mixtral 8x22B Instruct V0.1Ollamafreefree66K
24Mixtral 8x7B Instruct V0.1Ollamafreefree33K
25Qwen3 Coder:480b CloudOllamafreefree262K
26AutoOpenRouterfreefree2M
27FreeOpenRouterfreefree200K
28Nemotron 3.5 Lightning:freeOpenRouterfreefree1M
29Laguna S 2.1:freeOpenRouterfreefree262K
30Laguna Xs 2.1:freeOpenRouterfreefree262K
31Glm 5.2:freeOpenRouterfreefree256K
32Nemotron 3 Ultra 550b A55b:freeOpenRouterfreefree1M
33Minimax M3:freeOpenRouterfreefree1.0M
34Nemotron 3 Nano Omni 30b A3b Reasoning:freeOpenRouterfreefree256K
35Gemma 4 26b A4b It:freeOpenRouterfreefree262K
36Gemma 4 31b It:freeOpenRouterfreefree262K
37Minimax M2.7:freeOpenRouterfreefree197K
38Nemotron 3 Super 120b A12b:freeOpenRouterfreefree262K
39Llama 3.3 70B Instruct Turbo FreeTogether AIfreefree
40Qwen2.5 Coder 7BNebius$0.01$0.0333K
41Llama 3.2 3B InstructDeepInfra$0.02$0.02131K
42Mistral Nemo Instruct 2407DeepInfra$0.019$0.03131K
43Mistral NemoOpenRouter$0.019$0.03131K
44Meta Llama 3.1 8B Instruct TurboDeepInfra$0.02$0.04131K
45Meta Llama 3.1 8B InstructNebius$0.02$0.06128K
46Qwen3 4b Fp8Novita AI$0.03$0.03128K
47Meta Llama 3.1 8B InstructDeepInfra$0.03$0.05131K
48Llama 3.2 3b InstructNovita AI$0.03$0.0533K
49Meta Llama 3 8B InstructDeepInfra$0.03$0.068K
50Gemma 4 E4B ItDeepInfra$0.02$0.1131K
51Granite 4.0 H MicroCloudflare Workers AI$0.017$0.112131K
52L3 8B Stheno V3.2Novita AI$0.05$0.058K
53Nemotron 3.5 Lightning 30b A3bPerplexity$0.011$0.17$0.0011
54GPT Oss 20bOpenRouter$0.03$0.13131K
55Qwen3.7 FlashOpenRouter$0.03$0.13$0.0061M
56GPT Oss 20bDeepInfra$0.03$0.14131K
57Mistral Small 24B Instruct 2501DeepInfra$0.05$0.0833K
58Llama 3.1 8b InstantGroq$0.05$0.08131K
59Gemma 7b ItGroq$0.05$0.088K
60Llama 3.1 8b InstructOpenRouter$0.05$0.08$0.025131K

Other use cases

Frequently asked

Which model is cheapest for tool calling and agents?

Gemini Exp 1114 from Google Gemini currently leads this list, at $0/1M input tokens. Rankings are recomputed from the raw dataset on every refresh, so they track price cuts automatically.

How are these rankings calculated?

Each use case applies a capability filter (context size, vision, tool calling, cache pricing) and then ranks the surviving models by the blended cost that profile actually pays. See the methodology page for the exact formulas.

Is the cheapest model always the right choice?

No. Check the capability flags and context window, and benchmark quality on your own data — the cheapest model that fails your evals costs more than the second cheapest that passes them.