# Voice Prices

> Open pricing data for voice-agent APIs: STT, LLM, TTS, S2S, and VAD.

## Docs

- [Add your provider](https://prices.voicegateway.dev/contribute.md): Fix a price or add a new provider to voice-prices.
- [How fresh are these prices?](https://prices.voicegateway.dev/how-fresh.md): What verified, stale, imported, and seed mean, and who is allowed to set them.
- [Why Voice Prices](https://prices.voicegateway.dev/index.md): Open pricing data and cost calculation for voice-agent APIs, direct and through a gateway.
- [LLM pricing](https://prices.voicegateway.dev/llm/all-models.md): 1095 LLM model(s) from 31 provider(s), largest catalogs first.
- [Anthropic LLM pricing](https://prices.voicegateway.dev/llm/anthropic.md): 18 LLM model(s) from Anthropic, with input and output token rates.
- [Avian LLM pricing](https://prices.voicegateway.dev/llm/avian.md): 4 LLM model(s) from Avian, with input and output token rates.
- [AWS Bedrock LLM pricing](https://prices.voicegateway.dev/llm/aws.md): 69 LLM model(s) from AWS Bedrock, with input and output token rates.
- [Microsoft Azure LLM pricing](https://prices.voicegateway.dev/llm/azure.md): 18 LLM model(s) from Microsoft Azure, with input and output token rates.
- [Cerebras LLM pricing](https://prices.voicegateway.dev/llm/cerebras.md): 4 LLM model(s) from Cerebras, with input and output token rates.
- [Cohere LLM pricing](https://prices.voicegateway.dev/llm/cohere.md): 6 LLM model(s) from Cohere, with input and output token rates.
- [Deepseek LLM pricing](https://prices.voicegateway.dev/llm/deepseek.md): 4 LLM model(s) from Deepseek, with input and output token rates.
- [Fireworks LLM pricing](https://prices.voicegateway.dev/llm/fireworks.md): 13 LLM model(s) from Fireworks, with input and output token rates.
- [Google LLM pricing](https://prices.voicegateway.dev/llm/google.md): 31 LLM model(s) from Google, with input and output token rates.
- [Groq LLM pricing](https://prices.voicegateway.dev/llm/groq.md): 29 LLM model(s) from Groq, with input and output token rates.
- [HuggingFace (cerebras) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-cerebras.md): 1 LLM model(s) from HuggingFace (cerebras), with input and output token rates.
- [HuggingFace (fireworks-ai) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-fireworks-ai.md): 3 LLM model(s) from HuggingFace (fireworks-ai), with input and output token rates.
- [HuggingFace (groq) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-groq.md): 5 LLM model(s) from HuggingFace (groq), with input and output token rates.
- [HuggingFace (hyperbolic) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-hyperbolic.md): 12 LLM model(s) from HuggingFace (hyperbolic), with input and output token rates.
- [HuggingFace (nebius) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-nebius.md): 26 LLM model(s) from HuggingFace (nebius), with input and output token rates.
- [HuggingFace (novita) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-novita.md): 61 LLM model(s) from HuggingFace (novita), with input and output token rates.
- [HuggingFace (nscale) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-nscale.md): 20 LLM model(s) from HuggingFace (nscale), with input and output token rates.
- [HuggingFace (ovhcloud) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-ovhcloud.md): 7 LLM model(s) from HuggingFace (ovhcloud), with input and output token rates.
- [HuggingFace (publicai) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-publicai.md): 8 LLM model(s) from HuggingFace (publicai), with input and output token rates.
- [HuggingFace (sambanova) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-sambanova.md): 8 LLM model(s) from HuggingFace (sambanova), with input and output token rates.
- [HuggingFace (together) LLM pricing](https://prices.voicegateway.dev/llm/huggingface-together.md): 23 LLM model(s) from HuggingFace (together), with input and output token rates.
- [LiveKit Inference LLM pricing](https://prices.voicegateway.dev/llm/livekit.md): 33 LLM model(s) from LiveKit Inference, with input and output token rates.
- [Mistral LLM pricing](https://prices.voicegateway.dev/llm/mistral.md): 18 LLM model(s) from Mistral, with input and output token rates.
- [MoonshotAi LLM pricing](https://prices.voicegateway.dev/llm/moonshotai.md): 9 LLM model(s) from MoonshotAi, with input and output token rates.
- [Novita LLM pricing](https://prices.voicegateway.dev/llm/novita.md): 34 LLM model(s) from Novita, with input and output token rates.
- [OpenAI LLM pricing](https://prices.voicegateway.dev/llm/openai.md): 63 LLM model(s) from OpenAI, with input and output token rates.
- [OpenRouter LLM pricing](https://prices.voicegateway.dev/llm/openrouter.md): 461 LLM model(s) from OpenRouter, with input and output token rates.
- [OVHcloud AI Endpoints LLM pricing](https://prices.voicegateway.dev/llm/ovhcloud.md): 15 LLM model(s) from OVHcloud AI Endpoints, with input and output token rates.
- [Perplexity LLM pricing](https://prices.voicegateway.dev/llm/perplexity.md): 8 LLM model(s) from Perplexity, with input and output token rates.
- [Together AI LLM pricing](https://prices.voicegateway.dev/llm/together.md): 72 LLM model(s) from Together AI, with input and output token rates.
- [X AI LLM pricing](https://prices.voicegateway.dev/llm/x-ai.md): 12 LLM model(s) from X AI, with input and output token rates.
- [Serve a pricing feed](https://prices.voicegateway.dev/pricing-feed.md): Publish your rates once, in a file, so every tool that quotes you gets them right.
- [Speech-to-speech pricing](https://prices.voicegateway.dev/s2s/all-models.md): 6 S2S model(s) from 3 provider(s), cheapest first.
- [AWS Bedrock S2S pricing](https://prices.voicegateway.dev/s2s/aws.md): 1 S2S model(s) from AWS Bedrock, with the rate and where it came from.
- [Google S2S pricing](https://prices.voicegateway.dev/s2s/google.md): 1 S2S model(s) from Google, with the rate and where it came from.
- [OpenAI S2S pricing](https://prices.voicegateway.dev/s2s/openai.md): 4 S2S model(s) from OpenAI, with the rate and where it came from.
- [Speech-to-text pricing](https://prices.voicegateway.dev/stt/all-models.md): 47 STT model(s) from 7 provider(s), cheapest first.
- [AssemblyAI STT pricing](https://prices.voicegateway.dev/stt/assemblyai.md): 1 STT model(s) from AssemblyAI, with the rate and where it came from.
- [Deepgram STT pricing](https://prices.voicegateway.dev/stt/deepgram.md): 8 STT model(s) from Deepgram, with the rate and where it came from.
- [Groq STT pricing](https://prices.voicegateway.dev/stt/groq.md): 1 STT model(s) from Groq, with the rate and where it came from.
- [LiveKit Inference STT pricing](https://prices.voicegateway.dev/stt/livekit.md): 20 STT model(s) from LiveKit Inference, with the rate and where it came from.
- [LiveKit Inference (Scale) STT pricing](https://prices.voicegateway.dev/stt/livekit-scale.md): 13 STT model(s) from LiveKit Inference (Scale), with the rate and where it came from.
- [OpenAI STT pricing](https://prices.voicegateway.dev/stt/openai.md): 3 STT model(s) from OpenAI, with the rate and where it came from.
- [Telnyx STT pricing](https://prices.voicegateway.dev/stt/telnyx.md): 1 STT model(s) from Telnyx, with the rate and where it came from.
- [Text-to-speech pricing](https://prices.voicegateway.dev/tts/all-models.md): 63 TTS model(s) from 7 provider(s), cheapest first.
- [Cartesia TTS pricing](https://prices.voicegateway.dev/tts/cartesia.md): 1 TTS model(s) from Cartesia, with the rate and where it came from.
- [Deepgram TTS pricing](https://prices.voicegateway.dev/tts/deepgram.md): 2 TTS model(s) from Deepgram, with the rate and where it came from.
- [ElevenLabs TTS pricing](https://prices.voicegateway.dev/tts/elevenlabs.md): 5 TTS model(s) from ElevenLabs, with the rate and where it came from.
- [LiveKit Inference TTS pricing](https://prices.voicegateway.dev/tts/livekit.md): 25 TTS model(s) from LiveKit Inference, with the rate and where it came from.
- [LiveKit Inference (Scale) TTS pricing](https://prices.voicegateway.dev/tts/livekit-scale.md): 23 TTS model(s) from LiveKit Inference (Scale), with the rate and where it came from.
- [OpenAI TTS pricing](https://prices.voicegateway.dev/tts/openai.md): 3 TTS model(s) from OpenAI, with the rate and where it came from.
- [Telnyx TTS pricing](https://prices.voicegateway.dev/tts/telnyx.md): 4 TTS model(s) from Telnyx, with the rate and where it came from.
- [ai-coustics VAD pricing](https://prices.voicegateway.dev/vad/ai-coustics.md): 2 VAD model(s) from ai-coustics, with the rate and where it came from.
- [Voice activity detection pricing](https://prices.voicegateway.dev/vad/all-models.md): 2 VAD model(s) from 1 provider(s), cheapest first.