Skip to content
Sovyron

Cheapest Audio Model APIs

As of Nov 22, 2026, the cheapest audio model API is Phi-4-multimodal at $0.07 per 1M input and $0.11 per 1M output tokens on NanoGPT.

The table ranks the 20 cheapest of 51 qualifying models offered by at least 3 providers, by the cheapest offer's 3:1 input:output blended price. Models that accept or produce audio.

Cheapest Audio Model APIs, USD per 1M tokens, as of Nov 22, 2026
#Model
1Phi-4-multimodal$0.07NanoGPT$0.11NanoGPT$0.08128k3
2Doubao Seed 2.0 Mini$0.029Qiniu$0.28ZenMux$0.0925256k3
3Muse Spark 1.3 Contributor$0.1Bothub$0.2Bothub$0.1251.0M10
4Muse Spark 1.2 Contributor$0.1CommandCode$0.2CommandCode$0.1251.0M9
5MiMo-V2.6-Flash$0.1Vultr$0.25Vultr$0.13751.0M25
6Voxtral Small 24B 2507$0.1Eden AI$0.3Eden AI$0.1533k4
7Voxtral Small$0.1LLMTR$0.3LLMTR$0.1533k3
8Gemini 3 Flash Preview$0.07QiHang$0.43QiHang$0.161.0M28
9Gemini 2.5 Flash-Lite$0.095Jiekou.AI$0.38Jiekou.AI$0.16631.0M15
10MiMo-V2.5$0.14Impossibl$0.28Impossibl$0.1751.0M10
11Qwen3.8 Omni Flash$0.1126AIHubMix$0.38AIHubMix$0.17951.0M12
12Gemma 4 31B Instruct$0.12Venice AI$0.36Venice AI$0.18256k3
13Doubao Seed 2.0 Lite$0.087Qiniu$0.51ZenMux$0.195256k3
14Muse Ember$0.1Cloudflare AI Gateway$0.5Cloudflare AI Gateway$0.21.0M7
15Muse Ember 1.0$0.1DevPass (LLM Gateway)$0.5DevPass (LLM Gateway)$0.21.0M3
16Gemini 2.5 Flash$0.09QiHang$0.71QiHang$0.2451.0M24
17X-Ai/Grok-4-Fast-Non-Reasoning$0.2302.AI$0.5302.AI$0.2752.0M5
18Gemini 3.1 Flash Lite$0.125Kilo Gateway$0.75Kilo Gateway$0.28131.0M29
19Gemini 3.5 Flash$0.1857UnoRouter$1.1142UnoRouter$0.41781.0M33
20MiMo-V2.6-Pro$0.4Vultr$0.8Vultr$0.51.0M25

FAQ

What is the cheapest audio model API?

As of Nov 22, 2026, the cheapest audio model API is Phi-4-multimodal at $0.07 per 1M input and $0.11 per 1M output tokens on NanoGPT.

Which audio model APIs are cheapest after Phi-4-multimodal?

Doubao Seed 2.0 Mini ($0.0925 blended per 1M tokens on ZenMux), Muse Spark 1.3 Contributor ($0.125 blended per 1M tokens on Bothub), Muse Spark 1.2 Contributor ($0.125 blended per 1M tokens on CommandCode), as of Nov 22, 2026.

How is this list ranked?

By each model's cheapest metered offer, blended 3:1: (3 × input + output) ÷ 4 per 1M tokens. Free tiers and subscription plans are excluded, and a model needs at least 3 paid providers to qualify.

Other cheapest lists