February 2026
Model
NGen-4
A frontier general-purpose model for reasoning, coding, multimodal understanding, and production AI systems.
Overview
NGen-4 is the general frontier model tier for broad intelligence workloads, bringing together strong reasoning, coding, agentic execution, multilingual understanding, and multimodal performance in a single benchmark profile.
Context & Specs
Model Family
NGen 4 frontier series
Input support
Text, image, audio, video, and tool-oriented workflows
Best For
General frontier workloads, reasoning products, code assistants, and multimodal systems.
System Card
Available as a linked PDF for deeper model and safety details.
Key Capabilities
- Strong general intelligence across reasoning, coding, multilingual, and multimodal tasks.
- Designed for production API workflows and developer-facing applications.
- Benchmark profile spans knowledge, long context, engineering, agents, and vision.
- Built as the central model page for the NGen 4 release family.
Selected release
NGen-4
1 internal variant
Reasoning
Highest
Speed
Low
Input price
Rs 48 / 1M input tokens
Output price
Rs 88 / 1M output tokens
Input
Text, image, audio, video
Output
Text, image, audio, video
Overview
Model capabilities, context, and selection details.
NGen-4 has a 256K context window with a Low speed tier. Reasoning tier: Highest. Supported reasoning levels: Low, Medium, High. Input support: text, image, audio, video. Output support: text, image, audio, video.
- Context window: 256K (262,144 tokens).
- Speed tier: Low.
- Reasoning tier: Highest.
- Input support: text, image, audio, video. Output support: text, image, audio, video.
- Reasoning levels supported: Low, Medium, High.
Context window
256K (262,144 tokens)
Max output
64K (65,536 tokens)
Knowledge cutoff
Jun 2024
Rate limit
1,800 RPM
Pricing
Pricing is based on the selected internal release.
Input
Rs 48 / 1M input tokens
Cached input
Rs 12 / 1M cached input tokens
Output
Rs 88 / 1M output tokens
Intelligence vs Price Analysis
Top-tier models breaking into the Frontier Zone.
Model Evaluations
NGen-4 Evaluations
ShadCN benchmark charts for the new NGen-4 model using the benchmark sheet you provided.
NGen-4 Knowledge & Reasoning
Academic knowledge, Chinese evaluation, and graduate-level reasoning benchmarks.
NGen-4 Instruction Following
Instruction alignment, prompt following, and multi-turn challenge compliance.
NGen-4 Long Context
Long-context retrieval and sustained reasoning benchmarks.
NGen-4 Engineering & Coding
Software engineering, terminal, full-stack, and code benchmark performance.
NGen-4 General Agent
Tool use, browsing, planning, and agent benchmark performance.
NGen-4 Multilingual
Multilingual reasoning, translation, and cross-lingual problem-solving coverage.
NGen-4 Vision & Multimodal
Visual reasoning, OCR, math vision, and video understanding benchmarks.
Build with TNSA models
Explore model families, snapshots, and developer tools for production AI systems.