Price per 1M logo

Cheapest model for a customer-support bot

Blended cost estimate for a support bot profile: short prompts, medium answers, high volume, heavy prompt reuse.

Current leader: Gemini Exp 1114 (Google Gemini). Refreshed 2026-10-28 12:57 UTC.

#ModelProviderInput $/1MOutput $/1MCached $/1MContext
1Gemini Exp 1114Google Geminifreefree—1.0M
2Gemini Exp 1206Google Geminifreefree—2.1M
3Gemma 3 27b ItGoogle Geminifreefree—131K
4Gemma 4 26b A4b ItGoogle Geminifreefree—262K
5Gemma 4 31b ItGoogle Geminifreefree—262K
6Learnlm 1.5 Pro ExperimentalGoogle Geminifreefree—33K
7Labs Leanstral 1 5Mistral AIfreefree—262K
8Labs Leanstral 1 5 1Mistral AIfreefree—262K
9Gemma 2b It LoraCloudflare Workers AIfreefree—8K
10Mistral 7b Instruct V0.2 LoraCloudflare Workers AIfreefree—15K
11Llama 2 7b Chat Hf LoraCloudflare Workers AIfreefree—8K
12Gemma 7b It LoraCloudflare Workers AIfreefree—4K
13Codegeex4Ollamafreefree—33K
14CodegemmaOllamafreefree—8K
15CodellamaOllamafreefree—4K
16Deepseek Coder V2 BaseOllamafreefree—8K
17Deepseek Coder V2 InstructOllamafreefree—33K
18Deepseek Coder V2 Lite BaseOllamafreefree—8K
19Deepseek Coder V2 Lite InstructOllamafreefree—33K
20Deepseek V3.1:671b CloudOllamafreefree—164K
21GPT Oss:120b CloudOllamafreefree—131K
22GPT Oss:20b CloudOllamafreefree—131K
23Internlm2 5 20b ChatOllamafreefree—33K
24Llama2Ollamafreefree—4K
25Llama2 UncensoredOllamafreefree—4K
26Llama2:13bOllamafreefree—4K
27Llama2:70bOllamafreefree—4K
28Llama2:7bOllamafreefree—4K
29Llama3Ollamafreefree—8K
30Llama3.1Ollamafreefree—8K
31Llama3:70bOllamafreefree—8K
32Llama3:8bOllamafreefree—8K
33MistralOllamafreefree—8K
34Mistral 7B Instruct V0.1Ollamafreefree—8K
35Mistral 7B Instruct V0.2Ollamafreefree—33K
36Mistral Large Instruct 2407Ollamafreefree—66K
37Mixtral 8x22B Instruct V0.1Ollamafreefree—66K
38Mixtral 8x7B Instruct V0.1Ollamafreefree—33K
39Orca MiniOllamafreefree—4K
40Qwen3 Coder:480b CloudOllamafreefree—262K
41VicunaOllamafreefree—2K
42Ternary Bonsai 27BTogether AIfreefree—262K
43Qwen2.5 Coder 7BNebius$0.01$0.03—33K
44Qwen2.5 Coder 3B InstructNscale$0.01$0.03——
45Qwen2.5 Coder 7B InstructNscale$0.01$0.03——
46Llama 3.2 3B InstructDeepInfra$0.02$0.02—131K
47Paddleocr VlNovita AI$0.02$0.02—16K
48Llama 3.2 1b InstructNovita AI$0.02$0.02—131K
49Qwen2 1.5B InstructTogether AI$0.02$0.02—33K
50Mistral Nemo Instruct 2407DeepInfra$0.019$0.03—131K
51Meta Llama 3.1 8B Instruct TurboDeepInfra$0.02$0.04—131K
52DeepSeek R1 Distill Llama 8BNscale$0.025$0.025——
53Llama 3.1 8b InstructNovita AI$0.02$0.05—16K
54Ministral 3b 2512Mistral AI$0.1$0.1$0.01131K
55Ministral 3bMistral AI$0.1$0.1$0.01131K
56Ministral 3 3b 2512Mistral AI$0.1$0.1$0.01131K
57Llama Guard 3 8BNebius$0.02$0.06—128K
58Meta Llama 3.1 8B InstructNebius$0.02$0.06—128K
59Qwen2 VL 7B InstructNebius$0.02$0.06—131K
60Llama Prompt Guard 2 22mGroq$0.03$0.03—512

Other use cases

Cheapest chat APIs
Cheapest long-context models
Cheapest vision models
Cheapest tool-calling models
Cheapest reasoning models
Cheapest cached-input pricing
Cheapest embedding models
Cheapest model for summarisation at scale
Cheapest models for code generation

Frequently asked

› Which AI model should a support chatbot use?

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.