ChatMulti-purpose10B
Chat

GigaChat Lightning 10B

GigaChat · Apache 2.0

Mixture of Experts: 10B total parameters, 1.8B active per token.

Compact MoE model with 10B total parameters and around 1.8B active per token. Interesting efficiency play for chat workloads that want MoE-style speed without the memory cost of larger expert models.

Parameters
10B
Architecture
MoE (1.8B active)
Context
32,768 tokens
Released
2026-03-20
Engines
llama.cpp, ollama
Builder Tools
Continue, Open WebUI

Parameters

10B

VRAM

6 GB

Context

32K

Formats

2

GPUs

43

GigaChat Lightning 10B (10B) requires 6 GB VRAM at recommended quality (Q4_K_M). On NVIDIA RTX PRO 6000 Blackwell, expect approximately 325 tok/s at Q8_0. For the best experience, Starter AI Desktop ($582) is recommended.

Source: OwnRig methodology

VRAM (Recommended)

6 GB

Quantization

Q4_K_M

File Size

5.5 GB

Max Context

32K tokens

Primary Use

Chat

Memory

VRAM Requirements

QualityQuantizationVRAMFile Size
fullQ8_011 GB10 GB
recommendedQ4_K_M6 GB5.5 GB
Scaling

Context Length Impact

KV cache VRAM at Q4_K_M quality. Longer context = more memory.

ContextKV CacheTotal VRAM
2K102 MB6.1 GB
4K205 MB6.2 GB
8K410 MB6.4 GB
16K819 MB6.8 GB
32K1.6 GB7.6 GB

Compatible GPUs

43 devices
NVIDIA Grace Blackwell Ultra GB300Q8_0320 tok/sExcellent
Apple M4 Max (128GB Unified)Q8_072 tok/sExcellent
Apple M4 Max (36GB Unified)Q8_055 tok/sExcellent
Apple M4 Max (64GB Unified)Q8_061 tok/sExcellent
Apple M4 Pro (48GB)Q8_061 tok/sExcellent
Apple M4 Ultra (192GB)Q8_094 tok/sExcellent
AMD Radeon Pro W7900Q8_066 tok/sExcellent
NVIDIA RTX PRO 6000 BlackwellQ8_0325 tok/sExcellent
NVIDIA RTX PRO 6000 Blackwell Max-QQ8_0299 tok/sExcellent
Apple M1 Pro (16GB Unified)Q8_036 tok/sExcellent
Apple M2 Pro (16GB Unified)Q8_042 tok/sExcellent
Apple M4 Pro (24GB Unified)Q8_050 tok/sGood
NVIDIA GeForce RTX 3090Q8_066 tok/sGood
NVIDIA GeForce RTX 4090Q8_0110 tok/sGood
NVIDIA GeForce RTX 5090Q8_0143 tok/sGood
AMD Radeon RX 7900 XTXQ8_057 tok/sGood
AMD Radeon RX 9070Q8_096 tok/sGood
Apple M2 (16GB Unified)Q8_022 tok/sGood
Apple M3 (16GB Unified)Q8_024 tok/sGood
Apple M3 Pro (18GB Unified)Q8_038 tok/sAcceptable
Apple M4 (16GB Unified)Q8_044 tok/sAcceptable
NVIDIA GeForce RTX 3060 12GBQ4_K_M56 tok/sAcceptable
NVIDIA GeForce RTX 3080 10GBQ4_K_M72 tok/sAcceptable
NVIDIA GeForce RTX 4060 8GBQ4_K_M64 tok/sAcceptable
NVIDIA RTX 4060 Laptop (40-60W)Q4_K_M48 tok/sAcceptable
NVIDIA GeForce RTX 4060 Ti 16GBQ8_055 tok/sAcceptable
NVIDIA RTX 4070 Laptop (80-115W)Q4_K_M56 tok/sAcceptable
NVIDIA GeForce RTX 4070 SuperQ4_K_M96 tok/sAcceptable
NVIDIA GeForce RTX 4070 Ti 12GBQ4_K_M88 tok/sAcceptable
NVIDIA GeForce RTX 4070 Ti SuperQ8_072 tok/sAcceptable
NVIDIA RTX 4080 Laptop (120-150W)Q4_K_M80 tok/sAcceptable
NVIDIA GeForce RTX 4080 SuperQ8_082 tok/sAcceptable
NVIDIA RTX 4090 Laptop (150-175W)Q8_066 tok/sAcceptable
NVIDIA GeForce RTX 5080Q8_099 tok/sAcceptable
AMD Radeon RX 7600Q4_K_M50 tok/sAcceptable
Apple M1 (16GB Unified)Q8_012 tok/sAcceptable
AMD Radeon RX 9060 XT 16GBQ8_048 tok/sAcceptable
NVIDIA GeForce RTX 5060 8GBQ4_K_M74 tok/sAcceptable
NVIDIA GeForce RTX 5060 Ti 16GBQ8_062 tok/sAcceptable
Apple M1 (8GB Unified)Q8_0Not viable
Apple M2 (8GB Unified)Q8_0Not viable
Apple M3 (8GB Unified)Q8_0Not viable
AMD Radeon RX 9060 XT 8GBQ8_0Not viable

Showing 43 of 43 entries

Builder Context

GigaChat Lightning 10B is commonly used with Continue, Open WebUI.

FAQ

Frequently Asked Questions

How much VRAM does GigaChat Lightning 10B need?
GigaChat Lightning 10B requires 6 GB VRAM at recommended quality (Q4_K_M). At lower quality settings, it can fit in as little as 6 GB.
What is the best GPU for GigaChat Lightning 10B?
The NVIDIA RTX PRO 6000 Blackwell delivers the best performance for GigaChat Lightning 10B, achieving 325 tok/s at Q8_0 with an excellent rating.
Can I run GigaChat Lightning 10B on an RTX 4060 Ti?
Yes. On the NVIDIA GeForce RTX 4060 Ti 16GB, GigaChat Lightning 10B runs at 55 tok/s (Q8_0, acceptable).
What quantization should I use for GigaChat Lightning 10B?
For the best quality, use Q4_K_M (6 GB VRAM). If your GPU has limited VRAM, Q4_K_M (6 GB) is the most efficient option with acceptable quality.
All models

Data confidence: estimated. Source

VRAM requirements are calculated from model parameters and may vary by inference engine, context length, and batch size. Performance estimates are based on community benchmarks and should be verified for your specific configuration.GigaChat is a trademark of its respective owner. OwnRig is not affiliated with or endorsed by the model creator.