All models
Text / ReasoningAvailable via API80B total · 3B active

Vikasit 3 Coder

Code-specialized MoE. FIM, 262K context, agentic coding.

Overview

Vikasit 3 Coder is a code-specialized MoE — 80B total / 3B active, with fill-in-the-middle support, 262K context, and strong agentic-coding performance. Served live via the Vikasit API and used by the Vikasit CLI.

Specifications

Total parameters
80B total
Active parameters
3B active
Architecture
Hybrid MoE (Qwen3-Next base)
Context window
262K
Modalities
Text in → text out, fill-in-the-middle (FIM)
License
Apache 2.0

Capabilities

  • Agentic coding and multi-file edits
  • Fill-in-the-middle (FIM) completion
  • 262K context for whole-repo tasks
  • 80B quality at ~3B compute cost
Multilingual code + natural language.

Benchmarks

BenchmarkScore
MMLU-Pro80.5
GPQA74.5
AIME 202583.1
LiveCodeBench v658.9
SWE-bench Verified70.6
Aider-Polyglot66.2
HumanEvalN/A

Numbers from the Qwen3-Coder-Next Technical Report (arXiv:2603.00729). SWE-bench varies by harness (70.6 SWE-Agent / 71.3 OpenHands).

Hardware & deployment

PrecisionMemory
bf16~160 GB
INT4~45 GB

Quick start

Call Vikasit 3 Coder through the OpenAI-compatible Vikasit AI API at https://api.vikasit.ai/v1 using the model id vikasit-3-coder.

OpenAI-compatible Python (Vikasit AI API)
# pip install openai
import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.vikasit.ai/v1",
    api_key=os.environ["VIKASIT_API_KEY"],
)

resp = client.chat.completions.create(
    model="vikasit-3-coder",
    messages=[
        {"role": "user", "content": "Explain Vikasit 3 Coder in one sentence."}
    ],
)

print(resp.choices[0].message.content)
# or with curl
curl https://api.vikasit.ai/v1/chat/completions \
  -H "Authorization: Bearer $VIKASIT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "vikasit-3-coder",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Limitations

  • Full parameter set must be in memory (MoE)
  • Specialized for code; general chat below generalist models

Vikasit 3 Coder FAQ

How much does Vikasit 3 Coder cost?

Vikasit 3 Coder is served through the Vikasit AI API on usage-based, pay-as-you-go pricing billed per million input and output tokens — see the Vikasit AI pricing page for current rates. Because it is built on the open-weight Qwen3-Coder-Next 80B-A3B (Apache 2.0), you can also self-host the weights for free under the Apache 2.0 licence and pay only for your own compute.

Is Vikasit 3 Coder open weight?

Yes. Vikasit 3 Coder is built on Qwen3-Coder-Next 80B-A3B (Apache 2.0) and distributed under the Apache 2.0 licence, so the weights are openly available for self-hosting, fine-tuning, and commercial use, subject to the upstream licence terms.

How do I use Vikasit 3 Coder with the OpenAI SDK?

The Vikasit AI API is OpenAI-compatible. Point the OpenAI client's base URL at https://api.vikasit.ai/v1, set your Vikasit API key, and pass "vikasit-3-coder" as the model. The quick-start snippet above shows the exact Python call.

What context window does Vikasit 3 Coder support?

Vikasit 3 Coder supports a 262K context window. It is a 80B total (3B active) Hybrid MoE (Qwen3-Next base) model — full specifications are listed in the table above.

License & attribution

Apache 2.0

Built on Qwen3-Coder-Next 80B-A3B (Apache 2.0). Upstream copyright, license, and attribution notices are retained.

Qwen3-Coder-Next Technical Report