BestLLMfor EN Your hardware. Your LLM. Your call.
APIOpen data Find my LLM
Model fiche

Command R+ 104B (08-2024)

By Cohere · United States

Updated 2026-07-13

chat general multilingual
Parameters
104B
License
CC-BY-NC 4.0
Context
125k
VRAM (Q4)
60 GB
Released
August 2024 (refresh)

Overview

Cohere's 104B RAG and tool-use flagship from August 2024 — 128K context, 23 languages. Licensed CC-BY-NC, so non-commercial only without a Cohere agreement.

When to pick this model

  • You're building research or internal-only RAG systems
  • You need top-tier tool-use behavior in an open weights model
  • You need broad multilingual coverage across 23 languages
  • You're evaluating before signing a commercial agreement with Cohere

VRAM requirements by quantization

VRAM REQUIRED (GB)1624324880128Q4_K_M60 GBQ5_K_M72 GBQ8_0110 GBFP16208 GB
QuantizationVRAM required
Q4_K_M (recommended)60 GB
Q5_K_M72 GB
Q8_0110 GB
FP16 (no quantization)208 GB

VRAM figures include model weights plus a typical 8k KV cache and ~600 MB runtime overhead (Ollama / llama.cpp baseline). Add headroom for higher context lengths.

In practice, Command R+ 104B (08-2024) is server-class even at Q4_K_M (60 GB). Stepping up to Q8_0 nearly doubles the footprint to 110 GB, and unquantized FP16 weights take 208 GB — plan your GPU around the Q4 or Q5 figure unless you specifically need the higher fidelity.

Without a GPU, Command R+ 104B (08-2024) needs roughly 96 GB of system RAM to run on CPU via llama.cpp or Ollama — workable for background jobs, but far slower than GPU inference. Throughput estimates from our compatibility engine: around 1 tokens/sec on entry-level GPUs, on the order of 4 tokens/sec on a mid-range card, and up to 15 tokens/sec on high-end hardware — assuming the chosen quantization fully fits in VRAM.

What hardware do you need

The table below matches Command R+ 104B (08-2024) to common GPU memory tiers, using the highest-fidelity quantization that fully fits each card class. Spilling layers to system RAM works but costs most of the speed, so size your card to the quantization you actually want to run.

GPU memoryExample cardsBest fit for Command R+ 104B (08-2024)
8 GBRTX 5070 Laptop, RTX 5060, RTX 5060 Ti 8GBDoes not fit — needs 60 GB at Q4_K_M
12 GBRTX 5070, RTX 5070 Ti Laptop, RTX 4080 LaptopDoes not fit — needs 60 GB at Q4_K_M
16 GBRTX 5080, RTX 4080 Super, Radeon RX 9070 XTDoes not fit — needs 60 GB at Q4_K_M
24 GBRTX 4090, Radeon RX 7900 XTX, RTX 5090 LaptopDoes not fit — needs 60 GB at Q4_K_M
32 GBRTX 5090Does not fit — needs 60 GB at Q4_K_M

Which GPU should you buy to run Command R+ 104B (08-2024)?

To run Command R+ 104B (08-2024) locally at Q4, you need ~60 GB of VRAM. The best value for this is a Apple Mac Studio (64+ GB unified memory).

Check Apple Mac Studio price on Amazon →

As an Amazon Associate, BestLLMfor earns from qualifying purchases, at no extra cost to you. It does not influence our independent rankings.

Strengths

  • Best-in-class open RAG and tool-use at release
  • 128K context window
  • 23 language coverage
  • Higher throughput and lower latency than the April 2024 release

Limitations

  • CC-BY-NC 4.0 — no commercial use without a separate license
  • 60GB+ VRAM in Q4
  • Surpassed by newer 100B-class models on general benchmarks

Typical workloads

In our catalog grid, Command R+ 104B (08-2024) is filed under Advanced RAG, Tool use, Agents — the use cases where its size/quality trade-off makes the most sense. Its tags translate to concrete workloads: multilingual workloads.

The 125k-token context window covers long chats and mid-sized documents, though very large retrieval workloads will need chunking. It ships under the CC-BY-NC 4.0 license — commercial use is generally possible but read the specific terms before embedding it in a product.

Architecture & training

Architecture: Dense · optimized for RAG and tool-use · GQA

Training: 23 languages, +50% throughput / -25% latency vs April 2024 version.

Verdict

Strong RAG and tool-use, but the non-commercial license rules it out of most production deployments.

Quick start

ollama run command-r-plus:104b

Or use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.

Similar models worth comparing

Frequently asked questions

How much VRAM does Command R+ 104B (08-2024) need?

At the recommended Q4_K_M quantization, Command R+ 104B (08-2024) needs about 60 GB of VRAM. Q8_0 takes 110 GB, and unquantized FP16 weights take 208 GB.

Can Command R+ 104B (08-2024) run without a GPU?

Yes — with roughly 96 GB of system RAM it runs CPU-only through llama.cpp or Ollama. Expect a fraction of GPU speed, which is fine for background or batch jobs but slow for interactive chat.

What context window does Command R+ 104B (08-2024) support?

Command R+ 104B (08-2024) supports a 125k-token context window (128,000 tokens).

Can I use Command R+ 104B (08-2024) commercially?

Command R+ 104B (08-2024) ships under the CC-BY-NC 4.0 license. Commercial use is generally permitted subject to its terms — review the license text before shipping a product.

How fast is Command R+ 104B (08-2024) on consumer hardware?

Our compatibility engine estimates on the order of 4 tokens/sec on a mid-range GPU and up to 15 tokens/sec on high-end cards, assuming the quantization fully fits in VRAM.

Which quantization of Command R+ 104B (08-2024) should I download first?

Start with Q4_K_M (60 GB) — the standard size/quality sweet spot. Step up to Q5_K_M or Q8_0 only if you have VRAM headroom. It does not fit a single 24 GB consumer card — plan for multi-GPU or server hardware.

Tools

Is Command R+ 104B (08-2024) the right pick for you?

Compute self-hosted ROI → Back to catalog