LLM cost explorer

Estimate monthly spend from 100 to 1,000,000 calls on a log scale. Costs use each model's cheapest provider. Select a workload preset or customise input and output token counts per call.

Model Provider Monthly @ 10k calls Cost / call Input rate / 1M Output rate / 1M

Video cost explorer

Estimate monthly spend from 10 to 10,000 clips on a log scale. Costs use each model's cheapest tier and provider. Select a clip-duration preset; per-second models scale linearly with duration.

Model Provider Monthly @ 100 clips Cost / clip Billing Max duration

Audio cost explorer

Monthly cost estimates for TTS, STT, and music generation models. Costs are based on each model's listed API rate — credit-based models show the effective per-unit rate where available.

Text-to-speech (TTS)

1,000,000
Model Provider Rate Cost at 1M chars
GPT-Audio 1.5 OpenAI (Audio) ~$0.077 / min (est.) n/a — priced per minute
GPT-4o Audio Preview OpenAI (Audio) ~$0.096 / min (est.) n/a — priced per minute
Amazon Polly Amazon Web Services $4.00 / 1M chars $4.00
Azure Neural TTS Microsoft Azure $4.00 / 1M chars $4.00
Google Cloud TTS Google Cloud $4.00 / 1M chars $4.00
Fish Audio S1 Fish Audio $15.00 / 1M chars $15.00
Fish Audio S2 Pro Fish Audio $15.00 / 1M chars $15.00
Fish Audio S2.1 Pro Fish Audio $15.00 / 1M chars $15.00
OpenAI TTS-1 OpenAI (Audio) $15.00 / 1M chars $15.00
OpenAI TTS-1 HD OpenAI (Audio) $30.00 / 1M chars $30.00
Inworld TTS-1.5 Max Inworld AI $35.00 / 1M chars $35.00
ElevenLabs Flash v2.5 ElevenLabs $50.00 / 1M chars $50.00
Cartesia Sonic Cartesia $65.00 / 1M chars $65.00
ElevenLabs Eleven v3 ElevenLabs $100.00 / 1M chars $100.00
ElevenLabs Multilingual v2 ElevenLabs $100.00 / 1M chars $100.00

ElevenLabs is credit-based — rate shown is the pay-as-you-go character rate. Token-billed models (marked "est.") don't have a native per-character or per-minute price — their rate is an estimated $/minute-of-audio figure derived from OpenAI's published audio-token-to-duration ratio, not a verified price; see the model's own page for its real per-token pricing.

Speech-to-text (STT)

100 hours
Model Provider Rate Cost at 100 hrs
Whisper Large v3 Groq $0.0019 / min $11.10
AssemblyAI Universal-2 AssemblyAI $0.0025 / min $15.00
Deepgram Nova-3 Deepgram $0.0043 / min $25.80
Google Cloud STT Chirp 2 Google Cloud $0.0060 / min $36.00
OpenAI Whisper-1 OpenAI (Audio) $0.0060 / min $36.00
Azure Speech-to-Text Microsoft Azure $0.017 / min $100.20
Amazon Transcribe Amazon Web Services $0.024 / min $144.00

Deepgram Nova-3 has streaming and pre-recorded tiers; rate shown is the lower pre-recorded rate.

Music generation

100 songs
Model Provider Rate Cost at 100 songs
Udio Standard Udio $0.0042 / song $0.42
Suno v4 Suno $0.016 / song $1.60
MusicGen Large Meta AI $0.042 / song $4.20
Stable Audio 2.0 Stability AI $0.20 / song $20.00
Stable Audio 2.5 Stability AI $0.20 / song $20.00
Stable Audio 3.0 Stability AI $0.26 / song $26.00

Suno and Udio are consumer platforms; API access is not publicly available. Rates derived from subscription tiers (credits per song).

Cost explorer

Estimate monthly spend from 100 to 1,000,000 generations on a log scale. Costs use each model's cheapest tier, normalised to a per-image price (per-megapixel assumes 1 MP; per-second uses a model's typical generation time, or 3s). Subscription- and credit-only models aren't shown.

Model Provider Cost @ 1k Cost @ 10k Cost @ 100k