Nebius sells two products relevant to an API-pricing decision: Token Factory, an OpenAI-compatible inference API for open models (where each model is offered in a default and a -fast flavor with different latency and token pricing), and GPU instances billed per GPU-hour, from $2.15 for preemptible H100 up to $7.85 on-demand for HGX B300. The public pricing page documents instance, storage and network rates plus billing rules; Token Factory per-token prices are only shown after login. Commitment reservations are advertised at up to 35% below on-demand rates.

No free tier or USD sign-up credit is stated on official pages (checked 2026-09-16); the docs describe promo codes that grant free credits only when Nebius issues one in a special offer, and the first card payment is $25 credited to the account balance.

Novita AI runs 200+ models across text, image, audio and video behind an OpenAI-compatible API at api.novita.ai/openai, plus serverless endpoints, GPU instances and agent sandboxes, all billed per token or per GPU-second. Published model rates start at $0.02/1M input and $0.05/1M output for Llama 3.1 8B Instruct, with DeepSeek V4 Flash at $0.14/$0.28 on a 1M-token context, and three Ling 3.0 Flash variants (Fin, VL, Sante) listed at Free input and Free output with 256K context. Batch inference carries an introductory 50% discount on input and output tokens for supported models.

No standing free credit is stated on the current official pricing or docs pages (checked 2026-09-16); the pricing table lists Ling 3.0 Flash Fin/VL/Sante at Free input and output (256K context), and an official blog post from May 2025 advertised a limited-time $10 new-user credit that current pages do not repeat.

OVHcloud AI Endpoints is a serverless, OpenAI-compatible inference API for 40+ open-weight models, hosted in Gravelines, France, and billed per million tokens per model with input and output priced separately. Published catalogue rates include gpt-oss-20b at EUR 0.04 per 1M input and EUR 0.15 per 1M output, gpt-oss-120b at EUR 0.08/EUR 0.40, Meta-Llama-3.3-70B-Instruct at EUR 0.67/EUR 0.67 and Qwen3.5-397B-A17B at EUR 0.60/EUR 3.60, with embeddings at EUR 0.01 per 1M input tokens and whisper-large-v3 at EUR 0.00004083 per second of audio. Several models are listed as Free in the catalogue (Qwen3Guard-Gen-8B and Qwen3Guard-Gen-0.6B, stable-diffusion-xl-base-v10 and the nvr-tts voices), and a Batch tier in beta is documented at 50 percent savings for bulk processing.

No credit or allowance stated on official pages (checked 2026-09-16); the official AI Endpoints catalogue instead lists specific models as Free (Qwen3Guard-Gen-8B and -0.6B, stable-diffusion-xl-base-v10, and the nvr-tts-de-de/it-it/en-us/es-es voices) with all other models billed per million tokens.

OpenCode Zen is the model gateway from the OpenCode team that serves a curated, tested set of coding models. It includes a set of free models (Big Pickle, MiMo-V2.5 Free, Ling 3.0 Flash Fin Free, Nemotron 3 Ultra Free, Nemotron 3.5 Lightning Free, Muse Spark 1.3 Contributor Free) plus pay-as-you-go paid models, reachable at https://opencode.ai/zen/v1.

Free models are genuinely $0 per token (documented); no separate monthly USD credit.

OpenVecta provides an AI model API with an OpenAI-compatible endpoint. It focuses on simplicity and ease of use, with reports of a free tier or trial credits available for new users. Details were checked against the provider's own page at openvecta.com.

Free credits on signup for OpenAI-compatible inference across LLMs, embeddings, and reasoning models

Perplexity's API blends request and token pricing: Search API at $5 per 1,000 requests; Sonar at $1/$1 tokens plus a $5–$12 per-1K request fee that scales with search depth; and the new Agent API serving OpenAI, Anthropic, Google and xAI models at direct provider rates with no markup, plus per-invocation tool pricing (web_search $0.0025). Sonar is sunset after September 27, 2026 — the Agent API is the successor.

No free tier on the API (checked 2026-09-16). Consumer Pro/Max plan credits (35K bonus + 10K monthly) apply to the app, not the API.

StepFun's open platform exposes Step 3.7 Flash and Step 3.5 Flash reasoning models, a vision model and a full speech stack, priced per 1M tokens, per hour of audio or per 10,000 characters. Step 3.7 Flash is 1.35 CNY per 1M input tokens on a cache miss, 0.27 CNY cached input and 8.1 CNY output; Step 3.5 Flash is 0.7 / 0.14 / 2.1 CNY, and four audio preview models are marked limited-time free in the price table. Rate limits are tiered by cumulative top-up, from V0 with 5 concurrent requests, 10 RPM and 5,000,000 TPM at zero spend to V5 with 200,000 RPM at 10,000 CNY.

No numeric free grant is published on official pages (checked 2026-09-16); the billing doc confirms a gift-balance account is drawn down before paid balance but states no amount, while stepaudio-3-realtime-preview, stepaudio-3-chat-preview, stepaudio-3-gen-preview and stepaudio-3-music-preview are listed as limited-time free in the price table.

OpenTyphoon is a Thai-focused AI model platform by SCB 10X. It provides free access to models like Typhoon 2.5 via its hosted API at opentyphoon.ai. Production-grade access is offered through partners like Together AI and Float16.

No public free-tier information found in the source reference; treat free-tier availability as unverified rather than confirmed absent.

mimo.xiaomi.com is Xiaomi's MiMo model site, listing MiMo-V2.5-Pro, MiMo-V2.5, MiMo-V2.5-TTS and MiMo-V2.5-ASR alongside a web demo, API access, papers and blog posts. Paid access runs through the MiMo Platform Token Plan, priced in USD per month: Lite $5.28 (list $6), Standard $14.08 ($16), Pro $44 ($50) and Max $88 ($100), covering 49.2B to 984B credits with full V2.5 text, multimodal and speech model access and free TTS models as a limited-time offer. MiMo Code, Xiaomi's coding agent, is documented as startable with no login ('无需登录，开箱即用') and its provider doc points to an OpenAI-compatible endpoint at api.xiaomimimo.com/v1 with an Anthropic-compatible option at /anthropic.

No free credit grant stated: MiMo Code is documented as usable free without login and Token Plans include 'Free TTS Models - Limited time offer', but no USD credit amount is published (mimo.xiaomi.com/mimocode and platform.xiaomimimo.com/token-plan, checked 2026-09-16).

iFlyTek's Spark (星火) open platform sells its models through xinghuo.xfyun.cn in CNY per million tokens: the flagship Spark-X2.5 is listed at a limited-time 50% discount of 1.60 CNY per million input tokens, 6.00 CNY per million output tokens and 0.24 CNY per million cache-hit tokens. Spark X2 is quoted at 2–3 CNY per million tokens and Spark-X2-Flash at 1–2 CNY, while the small dense models Spark-X2.5-4B (limited-time free) and Spark-X2.5-1.7B (free) are offered at no cost. The same catalogue resells third-party open models such as GLM-5.2 (8 CNY in / 28 CNY out per million) and DeepSeek-V4-Flash (1 CNY in / 2 CNY out), alongside fine-tuning rates from 6 CNY per million tokens.

No cash credit: the platform states new users get 10,000 free voice interactions and 200,000 free tokens per Spark model ('每个模型20万tokens免费额度') on xinghuo.xfyun.cn/sparkapi; no CNY or USD credit figure is published (checked 2026-09-16).

spark-api-open.xf-yun.com is the OpenAI-compatible HTTP endpoint for iFlyTek's Spark (讯飞星火) LLM. It provides access to multimodal models via the iFlyTek Open Platform. A permanently free 'Lite' version is available for developers.

Spark Lite (model=lite) permanently free since May 2024: unlimited tokens, QPS caps only - unquantifiable; one-off token packs for other tiers (~2M tokens ~ $2, promo-dependent); real-name registration required.

xAI's API serves the Grok family — flagship grok-4.6 with 500K context at $2.00/$6.00 per 1M and configurable reasoning — plus Voice (agent/TTS/STT) and Imagine image/video generation, all in an OpenAI-compatible shape. X Search and Web Search are first-party server-side tools; free credits are promotional, not a published standing program.

No standing free-credit program published in xAI's docs (checked 2026-09-16). Promotional credits vary by account/region/campaign — confirm in the console at sign-up.