Models & APIs

Azure AI Speech

AppFree + paid

Reads text in a natural Turkish voice and turns speech into text.

What sets it apart: Turkish voices like Emel and Ahmet; a free quota every month.

For: Developers adding voiceover to their apps or video automations.

Cerebras Inference

AppFree + paid

An API service that runs open models very fast on its own large chips.

What sets it apart: Very high response speed; the trial tier is free.

For: Developers who want very fast model responses.

Cloudflare Workers AI

AppFree + paid

An API service that runs open models on Cloudflare's network.

What sets it apart: 10,000 Neurons free every day; your data isn't used for training.

For: Developers who host their site on Cloudflare.

Cohere API

AppFree + paid

Reach document search, ranking and chat models through an API.

What sets it apart: Strong search-result ranking (Rerank) and text embedding models.

For: Developers building search or Q&A over company documents.

Deepgram

AppFree + paid

A speech-to-text API; turns phone audio into text live or from recordings.

What sets it apart: Under half a cent a minute; Turkish works in the Nova-3 model, including live transcription.

For: Developers and agencies building voice agents and call transcription.

A small, open text-to-speech model that turns Turkish text into speech offline.

What sets it apart: About 34 MB; reads numbers, dates and money as written, and also runs on the CPU.

For: Developers building Turkish voice assistants, call centers or voiceovers.

fal.ai

AppPaid

Call image, video and audio models with one API key, pay as you go.

What sets it apart: Models like FLUX, Kling, Seedance and Nano Banana on one balance; no subscription.

For: People connecting image and video generation to code or an n8n flow.

FLUX

ModelFree + paid

Generates images from text and edits the images you give it with written instructions.

What sets it apart: With Kontext it changes the background and scene without distorting the product; the schnell version is open and free for commercial use.

For: People making product and ad images, and people connecting image generation to code.

Gemini TTS (Gemini API speech generation)

ModelPaid

Voice text in many languages, Turkish included, with the Gemini API.

What sets it apart: 30 preset voices, hundreds more; single- or two-speaker voiceover.

For: Developers adding voice to their apps via an API.

GLM (Z.ai)

AppFree + paid

A free chat site and cheap GLM models that connect to coding agents.

What sets it apart: A monthly coding plan for Claude Code, Cline and OpenCode; some models are free.

For: People looking for a cheaper model in their coding agent, people trying a free assistant.

Google AI Studio

AppFree + paid

Try Gemini models in the browser and get a free Gemini API key.

What sets it apart: Free API tier with no credit card; usable from Turkey.

For: People who want to prototype with Gemini or connect it to their app.

Google Cloud Text-to-Speech

AppFree + paid

Turns the text you write into natural-sounding audio files via an API.

What sets it apart: 30 Chirp 3 HD voices in Turkish; a free monthly character allowance per voice type.

For: People who want to add Turkish voiceover to video automations, product demos or apps.

Groq (GroqCloud)

AppFree + paid

An API service that runs open models very fast.

What sets it apart: Hundreds of tokens per second on its own chips; you start with a free plan.

For: Developers looking for a model that responds quickly.

Kimi (Moonshot AI)

AppFree + paid

A chat assistant that researches and prepares documents and sites.

What sets it apart: In Agent Swarm mode, it splits the work across sub-agents running at the same time.

For: People who want long research and multi-step work done with one command.

Meta Muse: API and Muse Code

ModelPaid

Call Meta's Muse models from the API, or have Muse Code write code in the terminal.

What sets it apart: OpenAI- and Anthropic-compatible endpoint; agents like Claude Code connect with just an address and a key. Muse Code is also available for a flat monthly fee.

For: Developers connecting models to their apps and people using coding agents in the terminal.

Mistral AI Studio (La Plateforme)

AppFree + paid

Reach the models of France's Mistral AI through an API.

What sets it apart: A free API mode with no credit card; Codestral as a code model.

For: Developers looking for a Europe-based model provider.

Modal

AppFree + paid

Run Python code and models in the cloud without setting up a server.

What sets it apart: You pay for the seconds you use; GPUs ready within minutes.

For: Developers who want to keep their automation or model running all the time.

NVIDIA API Catalog (build.nvidia.com)

AppFree

Try open models on NVIDIA's servers with a free API key.

What sets it apart: Models like GLM, Kimi, DeepSeek and Nemotron in one catalog; trying them is free.

For: Developers who want to try open models without setting up a server.

NVIDIA LocateAnything-3B

ModelFree

A model that finds the object you describe in a sentence and marks it with a box in an image.

What sets it apart: Finds objects, buttons or text from one sentence; small and fast.

For: People doing research on image analysis and screen automation.

OmniRoute

TerminalFree · open source

Routes your coding agents' requests through one address to the many model providers you connect.

What sets it apart: When one provider's quota runs out, the request automatically moves to another model.

For: Developers with several model keys who don't want to get stuck when a quota runs out.

OpenRouter

AppFree + paid

Reach hundreds of AI models with one API key; some are free.

What sets it apart: Models whose names end in ':free' are free; to switch models, you just change the name.

For: Developers connecting models to coding agents or automations.

Pinecone

AppFree + paid

A managed vector database that lets your chat assistant search your documents by meaning.

What sets it apart: Runs without a server; the Assistant upload-and-ask layer and an MCP endpoint come ready.

For: Teams that want their assistant to answer from their own documents without managing servers.

Qdrant

Open sourceFree · open source

An open-source vector database that can also run on your own server.

What sets it apart: Open source; installs with one Docker command and the data stays in-house; there's also a free cloud cluster.

For: Teams that want to keep their assistant's knowledge base on their own server.

Seedance (ByteDance)

ModelPaid

Generates short videos with sound from text, photos and reference videos.

What sets it apart: With many references, it can change the person, object and setting in a video.

For: People making ads and UGC videos.

Veo

ModelPaid

Generates video with sound from text and images; called from code and billed per second.

What sets it apart: Keeps the product and person consistent across scenes with 3 reference images; sound and dialogue come with the video.

For: People producing ad and social media videos through automation.