Inference API

Vikasit Inference

One OpenAI-compatible API for Vikasit's own models and the best of the open ecosystem. Pure pay-as-you-go per token — plus 2M free tokens a day on Vikasit Nova.

Vikasit Nova

2M tokens / day · free forever*

A capable general + coding model, free for every signed-in developer. 2 million tokens every day, no card required.

  • 2,000,000 tokens per day, free
  • OpenAI-compatible — drop-in base URL
  • Sign in with GitHub, get a key in seconds
  • Great for prototyping, scripts, and side projects

*Free daily allowance is subject to change. Fair-use limits apply per account.

Drop-in OpenAI-compatible

curl https://api.vikasit.ai/v1/chat/completions \
  -H "Authorization: Bearer $VIKASIT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "vikasit-nova",
    "messages": [{ "role": "user", "content": "Hello!" }]
  }'

Swap vikasit-nova for any model id below. Works with the OpenAI SDKs — just change the base URL and key.

Vikasit models(per 1M tokens)

Our own models, same rates as the Vikasit Code CLI.

ModelInputOutput
Vikasit 3
$0.21$0.30
Vikasit 3 Flash
$0.30$0.96
Vikasit 3 Fast
cache $0.39
$0.78$1.14
Vikasit 3 Max
cache $0.09
$0.81$2.85
Vikasit 3 Coder
cache $0.30
$1.20$4.80

Model catalog(per 1M tokens)

The wider ecosystem, resold through one key and one bill. DeepSeek, Qwen, Llama, Mistral, Gemma, GLM, Kimi, Claude, and more.

DeepSeek

ModelInputOutput
DeepSeek-V4-Flash
$0.10$0.20
DeepSeek-V3.2
$0.26$0.38
DeepSeek-V3.1-Terminus
$0.27$0.95
DeepSeek-R1-0528
$0.50$2.15
DeepSeek-V4-Pro
$1.30$2.60

Alibaba (Qwen)

ModelInputOutput
Qwen3-235B-A22B
$0.09$0.10
Qwen3.6-35B-A3B
$0.15$0.95
Qwen2.5-72B-Instruct
$0.36$0.40
Qwen3-Coder-480B-A35B
$0.40$1.60
Qwen3.5-397B-A17B
$0.45$3.00
Qwen3-Max
$1.20$6.00
Qwen3.7-Max
$2.50$7.50

Meta (Llama)

ModelInputOutput
Llama-4-Scout-17B
$0.10$0.30
Llama-3.3-70B-Turbo
$0.10$0.32
Llama-4-Maverick-17B
$0.15$0.60

OpenAI (open weights)

ModelInputOutput
gpt-oss-20b
$0.04$0.15
gpt-oss-120b
$0.05$0.45

Mistral

ModelInputOutput
Mistral-Nemo-Instruct
$0.02$0.04
Mistral-Small-3.2
$0.07$0.20

Google (Gemma)

ModelInputOutput
Gemma-3-27B-it
$0.08$0.16
Gemma-4-31B-it
$0.13$0.38

Zhipu (GLM)

ModelInputOutput
GLM-4.7-Flash
$0.06$0.40
GLM-5
$0.60$2.08

NVIDIA (Nemotron)

ModelInputOutput
Nemotron-3-Super-120B
$0.10$0.50
Nemotron-3-Ultra-550B
$0.50$2.50

Moonshot · MiniMax · Xiaomi · StepFun

ModelInputOutput
Step-3.5-Flash
$0.09$0.30
MiniMax-M2.5
$0.15$1.15
MiMo-V2.5
$0.40$2.00
Kimi-K2.5
$0.45$2.25
Kimi-K2.6
$0.75$3.50

Anthropic

ModelInputOutput
claude-sonnet-4-6
$3.00$15.00
claude-opus-4-8
$5.00$25.00

Building with the terminal instead? The Vikasit Code CLI is sold as a subscription plus pay-as-you-go for agentic coding.

Inference FAQ

Pure pay-as-you-go, per token, at the rates listed above. No subscription, no minimums. You only pay for what you call.

Yes. Point any OpenAI SDK at https://api.vikasit.ai with your Vikasit API key and set the model id. Chat completions, streaming, and tool calls all work.

Our free model. Every signed-in developer gets 2 million tokens per day at no cost. It is OpenAI-compatible like the paid models — ideal for prototyping. We may adjust or retire the free tier at any time.

The Vikasit Code CLI is sold as a subscription (Lite/Pro/Max) plus pay-as-you-go for agentic coding. Inference is the raw API for your own apps — pay-as-you-go only, with the free Vikasit Nova tier on top.