AI Battle > AI Gateway Comparison
150 AI gateways. Gateway rankings compare tokens per second, time-to-first-token and cost per million tokens for the same underlying models, so provider overhead is visible instead of hidden.
| Gateway | Category | Models | Starting price | Free tier | Best for |
|---|---|---|---|---|---|
| OpenRouter — One API, 640+ LLMs. Industry standard. | tokens | 640 | $10 credits | yes | Prototyping,Model switching,Fallback logic |
| AI/ML API — 400+ models, 0% markup | tokens | 441 | $20 top-up | yes | Cheapest GPT-5/Claude,High volume |
| Groq — 500+ tokens/sec, free tier forever | tokens | 20 | Free | yes | Realtime,Voice agents |
| Novita AI — Cheapest open-source LLMs | tokens | 200 | $0.15 / 1M | yes | DeepSeek,Qwen,Llama |
| Z.AI — GLM-5.3 coding plan for Claude Code, Cline and OpenCode. | tokens | 2 | From $18/mo | no | Coding agents,Claude Code,Flat monthly cost |
| fal.ai — 700+ image+video models, fastest inference | aggregator | 700 | Free credits | yes | All video models in one,Fast inference,One key |
| Kling AI — Best value, 66 daily free credits | video | 5 | Free 66 credits/day | yes | Free tier,Cinematic,Value |
| Higgsfield AI — Creator-friendly cinematic video + UGC | video | 30 | Basic $9/mo, Ultra $129/mo | yes | Cinematic AI,UGC ads,Product placement,Talking photo |
| Black Forest Labs Flux — Best open image model 2026 | image | 5 | $0.025/image | yes | Photorealistic,Open model,Cheap |
| ElevenLabs — Best voice cloning & TTS | audio | 50 | $5/mo Starter | yes | Voice cloning,TTS,Audiobooks,Dubbing |
| HeyGen — Best for sales & business video | avatar | 1100 | $24/mo Creator | yes | Sales,Training,Business explainer,Avatar |
| Together AI — Fast open-source inference + fine-tuning | tokens | 100 | $25 free credits | yes | RAG,Fine-tuning |
| DeepInfra — Cheap, no-frills Llama hosting | tokens | 50 | $5 deposit | no | Cheap Llama,Fallback |
| Fireworks AI — Sub-second Llama, best function calling | tokens | 80 | $1 free | yes | Function calling,Agents |
| OpenCode Zen — Curated for coding agents, 7 free models | tokens | 20 | $20 + $1.23 fee | yes | Claude Code,Cursor,Coding agents |
| Eden AI — One API for 200+ AI engines | aggregator | 200 | $10 free then $19/mo | yes | Multi-provider,No-code |
| Runway Gen-4 — Best creative suite, Sora alternative | video | 10 | $12/mo Standard | yes | Cinematic,Creative suite,Film |
| Google Veo 3.1 — Google's cinematic leader, native audio | video | 2 | Pay per use | yes | Cinematic with audio,Physics,Google ecosystem |
| ByteDance Seedance 2.0 — Best value/speed/quality balance | video | 2 | From $0.03/sec | yes | Value,Speed,UGC |
| Luma Dream Machine — Photorealistic, fast | video | 3 | $9.99/mo | yes | Photorealistic,Fast prototyping |
| Pika Labs — Best value at $8/mo | video | 4 | $8/mo | yes | Social video,Memes,Pikaffects,Cheap |
| MiniMax Hailuo 2.3 — Best character motion | video | 3 | $0.017/sec 512p | yes | Character motion,Consistency |
| Ideogram 3.0 — Best for text in image | image | 3 | $8/mo Basic | yes | Text in image,Logos,Typography |
| Midjourney — Most artistic, best aesthetics | image | 5 | $10/mo Basic | no | Artistic,Aesthetics,Concept art |
| Recraft V3 — Best for vector, icons, brand | image | 10 | Free tier | yes | Vector,Icons,Brand kits |
| Suno AI v4 — Best AI music generation | audio | 2 | $10/mo Pro | yes | Music,Songs with vocals,UGC music |
| Synthesia — Best for corporate video | avatar | 230 | $22/mo Starter | yes | Corporate,Training |
| Replicate — Run any open model in cloud | aggregator | 1000 | Free credits | yes | Open models,One API for everything |
| OpenAI — Frontier GPT models behind ChatGPT | tokens | — | $0.20 / 1M (GPT-5.6 Luna input) | no | Flagship reasoning,Multimodal apps,Agentic tool use,Voice and realtime |
| Anthropic — Claude models for agents and code | tokens | — | $1 / 1M (Claude Haiku 4.5 input) | yes | Coding agents,Long-context reasoning,Enterprise safety,Computer use |
| Google Gemini — Free-tier multimodal frontier API | tokens | — | $0.75 / 1M (Gemini 3.8 Flash input, through 2026-12-31) | yes | Free-tier prototyping,Video and multimodal,Long context,High-volume scale |
| xAI (SpaceXAI) — Grok 4.6 at $2 per million tokens | tokens | — | $2.00 / 1M (grok-4.6 input, <200k prompt) | no | Agentic tool calling,Real-time X/web search,500k-token context,Coding |
| Mistral AI — Open-weight models with EU sovereignty | tokens | — | $0.50 / 1M (Mistral Large input, per official FAQ) | yes | EU data sovereignty,Open-weight self-hosting,Document OCR,Speech and transcription |
| Cohere — Enterprise RAG, rerank and agents | tokens | — | $0.0375 / 1M (Command R7B input) | yes | Enterprise RAG,Reranking and retrieval,Private deployments,Document parsing |
| Perplexity — Search-grounded models and Agent API | tokens | — | $1 / 1M (Sonar input and output) | no | Web-grounded answers,Real-time research,Multi-model routing,Search and embeddings APIs |
| AI21 Labs — Open Jamba models and cost-routing gateway | tokens | 3 | $0.20 / 1M (Jamba Mini input) | yes | Long 256K context,Open-weight self-hosting,Private enterprise deploy,Cost-optimised routing |
| DeepSeek — Frontier reasoning tokens at rock-bottom prices | tokens | 2 | $0.15 / 1M (deepseek-flash input, off-peak) | no | Cheap reasoning,Long context,Code agents,High-volume batch |
| Moonshot AI (Kimi) — Kimi K3: 1M-context frontier at mid-tier prices | tokens | 4 | $0.95 / 1M (Kimi K2.6 input) | yes | Long-context agents,Coding,Multimodal chat,Chinese + English |
| Alibaba Cloud Model Studio (Qwen) — Qwen APIs from $0.03/1M across a vast catalog | tokens | — | $0.03 / 1M (qwen3.7-flash input) | yes | Cheap high-volume inference,Multimodal apps,Chinese-language products,Batch processing |
| Upstage — Solar Pro 4 promo rates plus best-in-class document AI | tokens | — | $0.09 / 1M (Solar Pro 4 promo through Oct 10, 2026) | yes | Korean and Japanese language,Document OCR and parsing,Enterprise / on-prem,Cost-efficient chat |
| Writer — Palmyra models plus an enterprise agent stack | tokens | 3 | $0.60 / 1M (Palmyra X5 input) | yes | Enterprise agents,Content and marketing,Governance and guardrails,Bring-your-own-model setups |
| Arcee AI — US open-weight Trinity models with a paid API | tokens | 10 | $0.25 / 1M (Trinity-Large-Thinking input) | yes | Open-weight agents,US-hosted inference,Self-host plus API hybrid,Open-source builders |
| Nous Research — Hermes models plus a credit-based model portal | tokens | — | From $20/mo (Plus, includes $22 monthly credits) | yes | Open-source models,Agent hosting,Multi-model routing,One bill for tools and models |
| Baidu Qianfan (ERNIE) — ERNIE 5.0 and DeepSeek on Baidu Cloud | tokens | — | ¥0.80 / 1M tokens | yes | Chinese-language apps,Enterprise agents,RAG with Baidu Search,ERNIE + third-party models |
| Tencent Hunyuan — Tencent's Hunyuan models and TokenHub gateway | tokens | — | ¥0.50 / 1M tokens | yes | Chinese enterprise apps,Hunyuan translation and role models,Multi-vendor gateway,Vision and embedding workloads |
| SiliconFlow — Open-model API cloud with free tiers | tokens | — | ¥0.70 / 1M tokens | yes | Open-source model hosting,Cost-sensitive production,Image, video and speech,Chinese and international users |
| StepFun — Multimodal agent models at flash prices | tokens | 20 | ¥0.70 / 1M tokens | yes | Agent workflows,Multimodal reasoning,Chinese-language apps,Low-cost flash inference |
| Sarvam AI — Indic-language LLM, speech and translation APIs | tokens | — | ₹29.28 / 1M tokens | yes | Indic-language voice agents,Indian language translation,Speech-to-text at scale,Document digitisation |
| Krutrim Cloud AI Studio — India-first cloud with per-token model APIs | tokens | 19 | ₹3.00 / 1M tokens | yes | India data-residency workloads,Indic-language apps,Open-model inference billed in INR,Combined GPU and inference billing |
| ModelScope — Alibaba's free open-model inference API | tokens | 48 | Free | yes | Prototyping open models,Free experimentation,Chinese open-source models,Comparing model outputs |
| Cerebras — Wafer-scale inference up to 30x faster than GPUs | tokens | 2 | $0.35 / 1M | yes | Ultra-low latency,Agents,Real-time chat,Reasoning |
| SambaNova — Fast open-model inference on the SambaCloud API | tokens | 7 | $0.22 / 1M | yes | Fast open models,Chatbots,Sovereign/on-prem AI,RAG |
| Nebius Token Factory — Open-model inference with transparent per-token pricing | tokens | 24 | $0.06 / 1M | yes | Open-weight models,Cheap batch inference,Embeddings,EU data residency |
| Baseten — Per-token Model APIs plus dedicated GPU deployments | tokens | 17 | $0.10 / 1M | yes | Production inference,Open-weight frontier models,Custom/fine-tuned models,Autoscaling |
| FriendliAI — Fast, cheap frontier open-weight inference | tokens | 8 | $0.14 / 1M | yes | Open-weight frontier models,Low-latency serving,Coding agents,Cost per token |
| Runpod — Serverless GPUs with public per-token endpoints | tokens | — | $0.50 / 1M | no | Serverless GPU,Cheap GPU pods,Inference endpoints,Fine-tuning |
| Portkey — Production AI gateway, now Prisma AIRS | tokens | 1600 | Free forever; Production from $49/mo | yes | Production reliability,Multi-provider routing,LLM governance,Guardrails |
| LiteLLM — Open-source AI gateway for platform teams | tokens | 1892 | Free (self-hosted, MIT) | yes | Self-hosting,Platform teams,Cost governance,Air-gapped deployments |
| Requesty — AI gateway with a flat 5% usage markup | tokens | 600 | Free tier (200 req/day); +5% on model cost | yes | EU data residency,Cost visibility,Model failover,Startups |
| Helicone — 0%-markup LLM gateway plus observability | tokens | 100 | Free (10k requests/mo); Pro $79/mo | yes | LLM observability,Debugging,Provider flexibility,Cost tracking |
| Vercel AI Gateway — Hundreds of models, one key, zero markup | tokens | 371 | Free tier; provider list price at 0% markup | yes | Zero-markup access,Coding agents,Multi-provider failover,Unified billing |
| Cloudflare AI Gateway — Free AI control plane for any provider | tokens | — | Free | yes | Free observability,Caching & rate limits,Provider fallback,Cloudflare-native apps |
| ZenMux — One API for 180+ top models, with AI insurance | tokens | 189 | $0.03 / 1M | yes | Production multi-model apps,Model routing,Cost tracing,Anthropic + OpenAI protocols |
| CometAPI — 500+ models at 20% off official rates | tokens | — | $0.12 / 1M | yes | Cheap multi-model access,App developers,Media generation,Cost consolidation |
| 302.AI — Pay-per-use hub of APIs and ready-made apps | tokens | — | $0.075 / 1M | no | Chinese + Western models,App prototyping,Image and video APIs,High concurrency |
| Venice AI — Private, uncensored inference via one key | tokens | 356 | $0.14 / 1M | yes | Privacy-sensitive apps,Uncensored chat,Crypto-funded usage,Multimodal generation |
| Nano-GPT — List-price API, crypto-friendly, no markup | tokens | 613 | $0.02 / 1M | yes | Crypto-funded usage,Cheap open models,OpenAI SDK drop-in,Subscription bundling |
| Amazon Bedrock — Multi-lab foundation models on AWS | tokens | 127 | $0.04 / 1M | yes | Enterprise workloads,Multi-model routing,RAG and agents,Compliance-heavy teams |
| Microsoft Foundry (Azure AI Foundry) — Microsoft's multi-model AI cloud | tokens | — | $0.40 / 1M | yes | Azure/enterprise shops,OpenAI models,Hybrid deployments,Data-residency needs |
| Vertex AI (Gemini Enterprise) — Google's Gemini-first model platform | tokens | 200 | $0.25 / 1M | yes | Gemini models,Google Cloud shops,Long-context work,Multimodal apps |
| OCI Generative AI — Oracle's enterprise GenAI inference cloud | tokens | — | $0.07 / 1M | yes | Oracle Cloud shops,Cost-sensitive inference,Data residency,Oracle Database apps |
| NVIDIA NIM — GPU-optimized inference microservices | tokens | — | Free (prototyping) | yes | Self-hosting open models,GPU optimization,On-prem and RTX,Agent prototyping |
| Cloudflare Workers AI — Serverless GPU inference at the edge | tokens | — | $0.011 / 1,000 Neurons | yes | Edge apps,Low-cost inference,Serverless teams,Prototyping |
| OVHcloud AI Endpoints — EU-sovereign serverless inference API | tokens | 40 | $0.05 / 1M | yes | EU data sovereignty,GDPR-sensitive apps,Budget chat and embeddings,OpenAI-compatible drop-in |
| DigitalOcean Gradient AI — One serverless API across many vendors | tokens | 58 | $0.05 / 1M | yes | One API for many vendors,Teams already on DigitalOcean,Model routing and evaluation,Mixing open and proprietary models |
| io.net (io.intelligence) — Free-tier inference on decentralized GPUs | tokens | 35 | $0.059 / 1M | yes | Free and low-cost experimentation,Open-weight models,Cost-sensitive batch workloads |
| CoreWeave Serverless Inference — Per-token serverless inference on an AI cloud | tokens | 31 | $0.03 / 1M | no | Enterprise per-token OSS serving,Teams wanting built-in tracing,Graduating to dedicated GPUs |
| Cursor Cloud Agents API — Programmatic cloud coding agents from Cursor | tokens | 48 | From $20/mo | yes | Cloud coding agents,CI/CD automation,Agent orchestration,Team agent workflows |
| Factory Droid — Agent API for enterprise software factories | tokens | 48 | From $20/mo | yes | Software factories,Enterprise SDLC automation,Cloud agent sessions,CI and code-review agents |
| Devin API — REST API for Cognition's autonomous AI engineer | tokens | — | From $20/mo | yes | Autonomous ticket-to-PR work,Parallel backlog burn-down,Code migrations and refactors,CI-triggered engineering tasks |
| Amp — Coding agent with an SDK for programmatic runs | tokens | — | From $20/mo | yes | Programmatic agent runs,Cloud agent automation,Multi-agent orchestration,Scripted coding tasks |
| Warp Agent API — Oz Agent API for cloud coding-agent runs | tokens | — | From $20/mo | yes | Cloud agent runs over REST,CI and backend automation,Self-hosted agent infrastructure,Multi-model agent orchestration |
| Ollama Cloud — Open-model cloud inference with a documented API | tokens | 20 | From $20/mo | yes | Cheap open-model inference,Privacy-sensitive apps,Local plus cloud hybrid setups,Coding-agent backends |
| Synthetic — Private open-model inference for $30/mo | tokens | 11 | $30/mo | no | Coding agents,Privacy-sensitive teams,Open-weight LLMs,Predictable flat-rate usage |
| Featherless — 40,000+ open models behind one API key | tokens | 21933 | $25/mo | no | Open-model variety,Roleplay and creative writing,Uncensored models,Self-serve agent use |
| Arli AI — Unlimited-token inference from $10/mo | tokens | — | From $10/mo | yes | Roleplay chat,Uncensored models,High-volume unlimited use,Coding agents |
| Chutes — Decentralized, TEE-secured serverless inference | tokens | 14 | $0.0245 / 1M | no | Decentralized inference,Confidential/TEE compute,Open-source models,Crypto (TAO) payments |
| Parasail — Serverless inference from $0.04 / 1M | tokens | 32 | $0.04 / 1M | no | Production inference,Cost-sensitive scale,Open and frontier models,Batch/offline jobs |
| Adobe Firefly — Commercially safe gen AI across Adobe apps | image | — | US$9.99/mo (Firefly Standard) | yes | Brand-safe commercial assets,Enterprise content ops,Adobe app integration,Batch production workflows |
| Bria AI — Licensed-data image AI built for enterprise IP safety | image | — | $0.018 per generation | yes | Enterprises needing IP indemnification,Commercially safe training data,Product shots and background removal at scale,On-prem or BYO-cloud deployment |
| ByteDance Seedream — ByteDance's flagship text-to-image family on ModelArk | image | — | $0.03/image | yes | 4K marketing and poster image generation,Chinese and English text rendering in images,Image editing and layer decomposition,Teams already running on BytePlus Cloud |
| Civitai — Model hub with a multi-provider generation API | image | — | $10/mo | no | Model discovery and community LoRAs,Running image and video workflows via one API,Buzz-funded hobby generation |
| Freepik (Magnific) — Freepik is now Magnific; API covers AI plus stock | image | — | PLN 50.50/mo (Premium, billed annually) | no | Apps needing AI generation plus stock imagery,Teams already paying for Magnific or Freepik plans,Marketing pipelines producing image and video |
| Google Gemini Image (Nano Banana / Imagen) — Nano Banana: cheap frontier image generation | image | — | $0.0336 per 1K image (Nano Banana 2 Lite); $0.039 per image on Gemini 2.5 Flash Image | yes | Character consistency,Conversational image editing,Text rendering in images,Cost-sensitive scale |
| ImageRouter — One API key, 350+ image, video and text models | image | 395 | $0.0002 per image | yes | Devs needing many image models, one key,Cheapest-provider auto routing,Fast prototyping on free models |
| Krea AI — Realtime image/video models with published API rates | image | 40 | $5/mo (Basic, billed annually) | yes | Devs wanting published per-second video rates,Realtime and interactive image generation apps,Brands training custom LoRA styles |
| Leonardo AI — Pay-as-you-go API for image, video and audio | image | — | $5 API credit | yes | Apps needing image, video and audio from one key,Teams leaving monthly API subscriptions for PAYG,Access to Veo, Kling and Nano Banana models |
| ModelsLab — Image, video, audio and 3D generation APIs | image | 10000 | $21/month | no | Cheap per-image generation APIs,Unlimited open-source model access,Replacing multiple model vendors |
| OpenAI Images (GPT Image) — GPT Image 2.5: frontier image generation | image | — | $0.005 per image (GPT Image 2, low quality, 1024x1536 / 1536x1024) | no | Photorealistic generation,Text inside images,Image editing,High-volume batches |
| Photoroom — Product-photo API from $0.02 per cutout | image | — | $20/mo for 1,000 images | yes | Ecommerce background removal at scale,Marketplace listing image pipelines,Ghost mannequin and virtual try-on,Product staging without a studio |
| Qwen-Image — Alibaba's Qwen image family with strong text rendering | image | — | $0.025/image | no | Chinese and multilingual text rendering in images,Infographics, posters and UI mockups,Image editing and image translation,Teams already on Alibaba Cloud |
| Runware — One API, 400K+ models, lowest-cost inference | image | — | $0.0006 per image | yes | Low-cost image and video inference,Many models behind one API key,Raw GPU compute billed by the second,EU or US data residency |
| Scenario — Custom-trained game asset generation platform | image | 170 | $0 free tier; paid from $15/mo | yes | Game studios training custom styles,Image, video and 3D asset pipelines,Teams needing an API-first workflow |
| SeaArt AI — All-in-one AI art community with a unified Vtrix API | image | — | $5.99/mo | no | Browser AI art with a huge model library,Credit-based image and video generation,Developers wanting one multimodal API key |
| Segmind — Pay-as-you-go gateway to image and video models | image | 200 | $0.0026/generation | no | Flux, Nano Banana and Seedream APIs,One key across 200+ models,Serverless image pipelines |
| Stability AI — Open-weight image models on a credit API | image | — | $0.025 per image (Stable Diffusion 3.5 Flash, 2.5 credits) | yes | Open-weight deployment,Self-hosting,Image editing and upscaling,Research and fine-tuning |
| TensorArt — Free model hosting plus a public workflow generation API | image | — | $1 day pass | no | Hosting and running community checkpoints and LoRAs,Workflow and ComfyUI generation via API,Low-cost image and video subscriptions |
| xAI Grok Imagine — Grok's Imagine models for image and video generation | image | — | $0.02/image | no | Cheapest per-image generation at $0.02,Short AI video clips at $0.05/sec,Apps built on OpenAI-compatible SDKs,Social and marketing creative workflows |
| Alibaba Wan — Wan 2.x and 3.0 video models with native audio | video | — | $0.02/sec | yes | High-volume text-to-video with audio,Image-to-video with first/last frame control,Regional data-residency deployments |
| Kie.ai — Low-cost unified API for Veo, Kling, Suno | video | — | $0.03/image (GPT Image 2 1K) | no | Cheap Veo and Kling video calls,Startups needing one media API,High-volume batch generation |
| LTX Studio / Lightricks LTX-2 — Open-weight 4K video model with synced audio | video | — | $0.03/sec | yes | Self-hosting video generation on own GPUs,Synchronized dialogue and ambience generation,Cheap 1080p text-to-video at scale |
| PiAPI — One API for Kling, Veo, Seedance and Sora | video | — | $0/month (Free plan) | yes | Kling, Veo and Sora video APIs,Multi-modal video, image and 3D,Pay-as-you-go from a credit balance |
| PixVerse — Effects-driven AI video for creators | video | — | $10 for 1,000 credits; docs cite about $1 per 5 videos (V6, 720p, 5s, no audio, Starter pack) | no | stylized effects and templates,anime and social video,high-volume 720p output,lip sync and avatar video |
| Pollo AI — Aggregator for 300+ video and image models | video | 300 | $15/mo (Lite, billed annually) | no | Developers comparing many video models,Marketing and short-form video,Teams wanting one API for many models |
| Sync.so — Studio-grade lipsync and visual dubbing API | video | — | $5/month | yes | Dubbing and localisation pipelines,Devs wanting lipsync SDKs,Creators testing lipsync cheaply |
| Tencent Hunyuan Video — Tencent's Hunyuan video generation via Cloud API | video | — | 1 yuan per credit (prepaid pack) | yes | Chinese-market apps on Tencent Cloud,Low-cost image and text to video,Teams wanting open-weight video models |
| Vidu — Cinematic video with real-time streaming | video | — | $0.005 per credit; from $0.02/sec (Vidu Q3-turbo 540P, off-peak) | no | cheap 540p drafts,reference-to-video consistency,real-time streaming avatars,long clips with native audio |
| WaveSpeedAI — Fast unified API for 1,000+ image and video models | video | 1000 | $0.005/image | yes | Cheap high-volume image generation,Per-second video model access,One key across 1,000+ models |
| AssemblyAI — Speech-to-text plus audio intelligence from $0.15/hr | audio | — | $0.15/hr | yes | Developers building production voice agents,Meeting and podcast transcription with diarization,Teams prototyping with a large free tier,Medical transcription using Medical Mode |
| CAMB.AI — Real-time dubbing and localization infrastructure | audio | 4 | $5/month (Essentials, 10K credits) | yes | Live sports and news dubbing,Multilingual media localization,Realtime voice agents |
| Cartesia — Realtime voice models with sub-90ms latency | audio | — | Free ($0/mo); paid plans from $5/mo | yes | Realtime voice agents,Low-latency TTS,Voice cloning,Streaming speech-to-text |
| Deepgram — Real-time STT, TTS and voice agents in one API | audio | — | $0.0043/min | yes | Voice agents needing low-latency STT plus TTS,High-volume batch transcription at $0.0043/min,Teams needing HIPAA and EU data residency,Self-hosted or on-premise voice AI deployments |
| Fish Audio — Open-source-grade TTS and voice cloning | audio | — | $0 (free tier); Plus $11/month | yes | Low-cost TTS at scale,Self-hosting open-weight models,Voice cloning and audiobooks |
| Gladia — Multilingual STT with 100+ languages and low latency | audio | — | $0.61/hr | yes | Multilingual meeting and contact-center transcription,EU teams needing GDPR and HIPAA compliance,Voice agents needing sub-300ms real-time STT,Developers preferring a prepaid pay-as-you-go wallet |
| Hume AI — Empathic EVI voice and Octave TTS | audio | — | $0/month Free plan; Starter $3/month | yes | Real-time empathic voice agents,Expressive TTS with acting instructions,Emotion and expression measurement,Voice cloning and custom personas |
| Loudly — Royalty-free AI music generation and API | audio | — | $0.125 per track generation (1,000 tracks/mo) | yes | Creators monetizing social media videos,Apps embedding royalty-free music,Podcast and YouTube background music |
| Murf AI — Studio TTS plus a low-latency streaming API | audio | 2 | $0.01 per 1,000 characters (Falcon) | yes | Realtime voice agents,eLearning and explainer voiceovers,Multilingual narration and dubbing |
| Resemble AI — Deepfake detection and voice cloning APIs | audio | — | $0/mo (Flex, pay-as-you-go credits) | no | Deepfake detection,Voice cloning,Audio watermarking,Identity verification |
| Rev AI — Accurate speech-to-text from $0.10 per hour | audio | — | $0.10/hr | yes | Budget batch transcription of English audio,Apps needing 57+ language coverage,Cheap Whisper access for voice AI apps,HIPAA-compliant enterprise deployments |
| Rime AI — Low-latency enterprise TTS from $0.03/1K chars | audio | — | $0.03 per 1K characters (~$0.03/min, Mist v3) | yes | Low-latency voice agents,Enterprise call-center TTS,On-prem or VPC voice deployment,Multilingual brand voice programs |
| Soniox — Realtime 60+ language STT from $0.10/hour | audio | — | $0.10/hour async STT; $0.12/hour real-time STT | no | Realtime multilingual transcription,Live speech translation in 60+ languages,Low-latency voice agents,High-volume async batch transcription |
| Speechify — Consumer TTS leader's per-character dev API | audio | 2 | $10/month Starter, then $10 per 1M characters | yes | English streaming TTS at scale,Accessibility and read-aloud apps,Teams avoiding credit maths |
| Speechmatics — Enterprise speech-to-text API in 55+ languages | audio | 3 | $0.129/hr | yes | Realtime transcription at scale,Multilingual and multi-speaker audio,On-prem and privacy-first deployments |
| Argil AI — AI avatars for studio-grade talking videos | avatar | — | $39/month | no | Marketer-facing avatar video at volume,Teams already on a paid plan,Script-to-video with b-roll and templates |
| Captions (Mirage) — AI video editing, avatars and the Mirage API | avatar | — | $24.99/mo Max plan | yes | Creators editing long footage fast,Social clips with 100+ language captions,Devs adding video captioning APIs |
| Colossyan — AI presenter video for training and enablement | avatar | — | Free (Starter); Professional $59/mo yearly | yes | Corporate training and onboarding video,Compliance and e-learning at scale,Teams that need SCORM and course authoring |
| Creatify AI — AI avatar video ads from any product URL | avatar | 100 | $39/mo (100 credits) | yes | Ecommerce and DTC ad teams,UGC-style avatar ads at volume,Agencies managing many brands |
| D-ID — Talking-head avatars and realtime agents | avatar | — | $4.70/mo (Studio Lite, annual billing) | yes | Talking-head avatar video at scale,Realtime streaming agents,Video translation and localization |
| Deepbrain AI — 2,000+ AI avatars with interactive avatar API | avatar | — | Free; Personal $24/mo | yes | Enterprise training and e-learning,Conversational avatars in finance and retail,Dubbing and localization at scale,Teams needing licensed stock avatars |
| Elai.io — Text-to-avatar video with public API on every plan | avatar | — | Free; Creator $23/mo annual, $29/mo monthly | yes | Fast explainer and training video,Localizing content into 75+ languages,Personalized video at scale via API |
| Hedra — Longform talking avatars, plus 98 models | avatar | 98 | $15/month for 1,500 credits; API avatar from 2.5 cents/sec (Character-3) | yes | talking and singing avatars,longform avatar video up to 10 minutes,multi-vendor model access,agencies and creators |
| Simli — Realtime lip-sync avatar API under 300 ms | avatar | — | Free; Hobby $10/mo | yes | Realtime voice and video agents,LiveKit and Pipecat pipelines,Sales, coaching and language-training bots,Developers needing low-latency lipsync |
| Tavus — Realtime conversational video avatars via API | avatar | — | Free developer tier; paid tiers from $22/mo | yes | Realtime conversational video agents,AI interviewers and SDR agents,Healthcare and L&D roleplay agents |
| Hyper3D Rodin (Deemos) — AI 3D assets from image, text or multi-view | 3d | — | $0 Free plan; paid from $30/mo | yes | Studios needing a documented 3D API,Game and AR asset pipelines,Quad topology and high-poly output |
| Immersity AI — 2D-to-3D depth, spatial video and immersive motion | 3d | — | $4.99/mo (Image, 500 credits) | yes | 2D photo to 3D depth maps,Spatial video and stereo conversion,Apple spatial photo and video export |
| Layer AI — Multi-model AI asset platform for game studios | 3d | 500 | $10/month for 300 CUs | no | Game studios shipping 2D and 3D assets,Teams wanting one API for all media,UA creative production at scale |
| Meshy — Text and image to 3D models with PBR texturing | 3d | — | $20/mo (Pro, 1,000 credits/mo) | yes | Game and XR asset pipelines,3D printing model generation,Automated text-to-3D at scale |
| Sloyd — Parametric and AI 3D asset generation | 3d | — | $0 Guest; API from $12 for 900 credits | yes | Game devs needing parametric templates,In-app user-generated 3D content,Roblox and 3D-printing asset pipelines |
| Tencent Hunyuan 3D — Text-to-3D API with cheap per-model credits | 3d | — | CNY 100 / 1,000 credits (CNY 0.10 per credit) | yes | Game and 3D asset pipelines,Bulk text-to-3D generation,Apps already on Tencent Cloud |
| Tripo AI — Credit-based 3D model generation API | 3d | — | $0.10 (text-to-3D, 10 credits) | yes | Indie game asset generation,Rigged character pipelines,Apps needing bulk 3D on a budget |