Updated

ElevenLabs alternatives

Of the 80 products launched in 2026 that we know the Speech-to-Text & Text-to-Speech of, 63% use ElevenLabs, against 50% of those launched in 2024. ElevenLabs is the most picked.

5 of the 9 are open source or self-hostable (Deepgram, Cartesia, AssemblyAI and 2 more). ElevenLabs also answers voice agents: those alternatives are below too.

5 open source or self-hostable0 with a free tier1 ready for AI agents

Not a ranking: each option is described by the situation it fits, with its trade-off.

What new products use instead

The share of makers' products we track that use each for Speech-to-Text & Text-to-Speech, by the year they launched on Product Hunt. The stack is the one we see today, not the one it launched with; a product can use more than one.

OptionLaunched 2024 · 42Launched 2025 · 65Launched 2026 · 80
ElevenLabs50%58%63%
Deepgram33%26%25%
Cartesia14%8%10%
AssemblyAI10%15%6%
Recall.ai10%8%6%
Fish Audio0%5%5%

ElevenLabs alternatives compared for Speech-to-Text & Text-to-Speech

How many makers' products we track use each for Speech-to-Text & Text-to-Speech.

OptionKindMakers' productsFree tierOpen source
ElevenLabsVoice generation135—Self-hostable
DeepgramTranscription65—Self-hostable
CartesiaVoice generation29—Self-hostable
AssemblyAITranscription26—Self-hostable
Recall.aiMeeting bots21—No
Fish AudioVoice generation9—No
OpenAI APIGeneral LLM provider1NoNo
WhisperOpen model you run0—Yes
KokoroOpen model you run0—Yes
Hume AIVoice generation0—No

ElevenLabs alternatives compared for voice agents

How many makers' products we track use each for voice agents.

OptionKindMakers' productsFree tierOpen source
ElevenLabsHosted platform0—Self-hostable
LiveKitOpen-source framework38—Yes
VapiHosted platform16—No
DailyReal-time infrastructure16—No
PipecatOpen-source framework5—Yes
Retell AIHosted platform4—No
OpenAI APISpeech-to-speech model0NoNo
LayercodeHosted platform1—No
BlandHosted platform0NoNo

When each fits: Speech-to-Text & Text-to-Speech

The whole decision →

Real-time or batch, and which direction. Streaming voice features need low latency above everything, while transcribing recordings is about accuracy and price per hour. Most providers lead in one direction — speech-to-text or text-to-speech — so many apps end up using two.

ElevenLabs (Voice generation): Natural, expressive voices and voice cloning for narration, content and characters. Trade-off: Credits run out fast on long audio, and cloning needs consent from the voice owner.

Deepgram

Transcription
Self-hostable

Best forStreaming speech-to-text with low latency, for live calls and voice interfaces.

Trade-offIts text-to-speech is built for voice agents, with a set of prebuilt voices and no voice cloning.

Used by 65 makers' products for this

Cartesia

Voice generation
Self-hostable

Best forText-to-speech with very low time to first audio, for live conversation.

Trade-offBuilt for real-time use, with fewer creator tools such as dubbing and sound effects.

Used by 29 makers' products for this

AssemblyAI

Transcription
Self-hostable

Best forAccurate transcription of recordings with speaker labels and extras like summaries and topic detection.

Trade-offFocused on speech-to-text; you need another vendor for speech output.

Used by 26 makers' products for this

Recall.ai

Meeting bots

Best forGetting recordings, transcripts and participant data from Zoom, Google Meet, Microsoft Teams and other calls through one bot API.

Trade-offBilled per recording hour on top of any transcription, and the bot is visible to everyone in the call.

Used by 21 makers' products for this

Fish Audio

Voice generation

Best forText-to-speech and quick voice cloning with a large library of community voices in 30+ languages.

Trade-offCommunity voices vary in quality and rights; its open-weight models are under the Fish Audio Research License, so check its terms before self-hosting commercially.

Used by 9 makers' products for this

OpenAI API

General LLM provider
From Pay per tokenllms.txtIn Lovable, Replit, Rork

Best forAdding transcription and speech output on the same account you already use for text.

Trade-offFewer voices and less tuning than specialist providers.

Used by 1 maker's product for this

Whisper

Open model you run
Open source

Best forTranscribing and translating audio in many languages with an open model on your own hardware.

Trade-offBatch-oriented, not built for real-time streaming; large models need a GPU for reasonable speed, and speaker labels need another tool.

No maker product tracked for this yet

Kokoro

Open model you run
Open source

Best forA small open-weight text-to-speech model that runs on a CPU or in the browser, under Apache 2.0.

Trade-offA fixed set of voices with no cloning, and less expressive than the leading hosted voices.

No maker product tracked for this yet

Hume AI

Voice generation

Best forExpressive text-to-speech (Octave) and an emotion-aware speech-to-speech interface (EVI).

Trade-offThe TTS and EVI APIs end on 2026-11-13, and account data is deleted after that.

No maker product tracked for this yet

When each fits: voice agents

The whole decision →

Hosted platform or open-source framework. Hosted platforms handle phone numbers, turn-taking and provider wiring, and charge per minute on top of model costs. Open-source frameworks give you full control over the pipeline and cost, but you run the real-time servers yourself. A speech-to-speech model skips the chain entirely: one model hears and answers, with less delay but less control over each step.

ElevenLabs (Hosted platform): Voice agents built on ElevenLabs voices, with turn-taking, tools, knowledge bases and phone numbers in one platform. Trade-off: Per-minute pricing plus LLM costs, and the best fit is with ElevenLabs' own speech models.

LiveKit

Open-source framework
Open source

Best forReal-time voice and video in your own app, with an agents framework on top of its WebRTC infrastructure.

Trade-offMore to learn and run than a hosted platform, unless you pay for LiveKit Cloud.

Used by 38 makers' products for this

Vapi

Hosted platform

Best forDevelopers who want to configure every part — models, voices, tools, phone numbers — through an API.

Trade-offPlatform fee per minute plus the pass-through model costs, which makes the total harder to predict.

Used by 16 makers' products for this

Daily

Real-time infrastructure

Best forWebRTC audio and video transport plus Pipecat Cloud, which hosts Pipecat agents so you don't run the real-time servers.

Trade-offHosted only, billed per minute, and you still build the agent logic yourself.

Used by 16 makers' products for this

Pipecat

Open-source framework
Open source

Best forA Python pipeline where you choose and swap every speech, LLM and voice component.

Trade-offYou host and scale the real-time processes yourself.

Used by 5 makers' products for this

Retell AI

Hosted platform

Best forPhone agents for bookings and customer service, set up largely through a visual builder.

Trade-offLess room for custom pipelines than building on an open framework.

Used by 4 makers' products for this

OpenAI API

Speech-to-speech model
From Pay per tokenllms.txtIn Lovable, Replit, Rork

Best forThe Realtime API, where one model listens and speaks directly, over WebRTC, WebSocket or SIP phone calls.

Trade-offLess control over each step than a chained pipeline, a small set of voices, and audio tokens cost more than text.

No maker product tracked for this yet

Layercode

Hosted platform

Best forLow-latency voice agents whose speech pipeline ran on Cloudflare's edge while you supplied the agent logic.

Trade-offRenamed Toyo and pivoted; layercode.com redirects to toyo.ai and the SDKs stopped in February 2026.

Used by 1 maker's product for this

Bland

Hosted platform
From $0.14/min

Best forPhone agents at one per-minute price that includes the LLM, transcription and voice, with on-prem or VPC deployment for enterprises.

Trade-offThe model, transcription and voices are bundled into its price rather than chosen per provider as on Vapi; the cheapest plan caps you at 100 calls a day.

No maker product tracked for this yet

Questions

What is the best alternative to ElevenLabs?

It depends on what you need it for; we don't rank them. Among products launched in 2026, the most picked for Speech-to-Text & Text-to-Speech instead is Deepgram (25%). You're building live conversation where every millisecond of delay shows: Cartesia and Deepgram. Users upload recordings to be transcribed and analysed: AssemblyAI and Deepgram. Your product needs recordings and transcripts from Zoom, Meet or Teams calls: Recall.ai.

Is there a free alternative to ElevenLabs?

None of the Speech-to-Text & Text-to-Speech alternatives we list has a free tier.

Is there an open-source alternative to ElevenLabs?

Yes. For Speech-to-Text & Text-to-Speech: Deepgram (self-hostable), Cartesia (self-hostable), AssemblyAI (self-hostable), Whisper and Kokoro.