Calculate the TPS of the GeForce GTX 680M on local AI models

NVIDIA 4 GB GDDR5 115 GB/s June 2012

Every model in our catalogue assessed against this card at the context length and minimum quality you choose. Speed is an estimate for a single request, calculated from this card's memory bandwidth and the size of each model once compressed.

Calculated for this card

97 models it can run

679 models in our catalogue altogether

Largest model it holds

DeciLM 6B

5.7B · Q3_K_M · 19.6 tok/s

Fastest model

Gemma 3 QAT 1B

41.5 tok/s · 1B

Which AI models can run on a GeForce GTX 680M?

Set the inputs, read the answer

More context means more memory for the conversation cache. Speed is for a fresh conversation and does not change with this setting.

Hides models that would only fit by being compressed below this point.

97 models match

Calculating
Quantisation Fit
41.5 tok/s

15–83 · low confidence

Gemma 3 1B 1B Mar 2025 1.8 GB 33k tokens Q8_0 Comfortable
41.5 tok/s

15–83 · low confidence

Gemma 3 QAT 1B 1B Apr 2025 1.8 GB 33k tokens Q8_0 Comfortable
41.5 tok/s

15–83 · low confidence

HGRN 1B (WT 103) ≈ 1B Nov 2023 1.8 GB 131k tokens ? Q8_0 Comfortable
41.5 tok/s

15–83 · low confidence

LLama 3..2 Typhoon 2 1B ≈ 1B Dec 2024 1.8 GB 131k tokens ? Q8_0 Comfortable
41.5 tok/s

15–83 · low confidence

OLMo-1B ≈ 1B Feb 2024 1.8 GB 131k tokens ? Q8_0 Comfortable
41.5 tok/s

15–83 · low confidence

Pythia-1b ≈ 1B Apr 2023 1.8 GB 131k tokens ? Q8_0 Comfortable
38.4 tok/s

13–77 · low confidence

OpenELM-1.1B ≈ 1.1B May 2024 1.9 GB 131k tokens ? Q8_0 Comfortable
37.7 tok/s

13–75 · low confidence

DeciCoder-1B ≈ 1.1B Aug 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
37.7 tok/s

13–75 · low confidence

SantaCoder ≈ 1.1B Jan 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
37.7 tok/s

13–75 · low confidence

TinyLlama-1.1B (1T token checkpoint) ≈ 1.1B Oct 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
37.7 tok/s

13–75 · low confidence

TinyLlama-1.1B (3T token checkpoint) ≈ 1.1B Oct 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
34.6 tok/s

12–69 · low confidence

EXAONE 4.0 (1.2B) ≈ 1.2B Jul 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
34.6 tok/s

12–69 · low confidence

MinerU2.5 ≈ 1.2B Sep 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
34.6 tok/s

12–69 · low confidence

Pleias 1.0 1.2B ≈ 1.2B Dec 2024 2.0 GB 131k tokens ? Q8_0 Comfortable
34.6 tok/s

12–69 · low confidence

Pleias-RAG-1B ≈ 1.2B Apr 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
33.7 tok/s

12–67 · low confidence

Llama 3.2 1B 1.2B Sep 2024 2.2 GB 54k tokens Q8_0 Comfortable
33.3 tok/s

12–66 · low confidence

MiniCPM-1.2B ≈ 1.2B Jun 2024 2.0 GB 131k tokens ? Q8_0 Comfortable
31.9 tok/s

11–64 · low confidence

DeepSeek Coder 1.3B ≈ 1.3B Jan 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
31.9 tok/s

11–64 · low confidence

DeepSeek-VL-1.3B ≈ 1.3B Mar 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
31.9 tok/s

11–64 · low confidence

DigiRL ≈ 1.3B Jun 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
31.9 tok/s

11–64 · low confidence

GLA Transformer 1.3B ≈ 1.3B Aug 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
31.9 tok/s

11–64 · low confidence

Janus 1.3B ≈ 1.3B Oct 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
31.9 tok/s

11–64 · low confidence

Kosmos-2.5 ≈ 1.3B Aug 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
31.9 tok/s

11–64 · low confidence

Otter ≈ 1.3B May 2023 2.1 GB 131k tokens ? Q8_0 Comfortable
31.9 tok/s

11–64 · low confidence

Phi-1 ≈ 1.3B Oct 2023 2.1 GB 131k tokens ? Q8_0 Comfortable

Speeds are estimates for a single request — one conversation at a time — calculated from memory bandwidth, model size and quantisation. Real throughput varies with the inference runtime and its version. Figures published by hardware vendors measure many simultaneous requests and are much higher.

On record

GeForce GTX 680M full specification

Everything on record for this board, ordered by how much it bears on running a language model rather than by how a spec sheet would list it. Memory comes first because it decides the outcome; the rest is context.

Memory

The two specifications that decide what this card can run and how quickly. Capacity sets which models fit; bandwidth sets how many tokens per second they produce once they do.

Memory size
4 GB
Memory bandwidth
115 GB/s
Memory type
GDDR5
Memory bus width
256 bit
Memory clock
900 MHz

The chip

Which processor is on the board and how it was manufactured. A smaller process size generally means more performance for the same power.

Graphics processor
GK104
Architecture
Kepler
Generation
GeForce 600M
Foundry
TSMC
Process size
28 nm
Transistors
3.5 billion
Transistor density
12,000 K/mm²
Die size
294 mm²
Package
BGA-1745
Released
4 June 2012

Clock speeds

How fast the processor runs. Worth far less here than on a gaming benchmark: generating text is limited by memory bandwidth, so a higher clock barely moves the result.

Base clock
719 MHz
Boost clock
758 MHz

Processing units

What the chip contains. These drive graphics performance and matter mainly for processing a long prompt rather than for producing the answer.

Shading units
1,344
Texture mapping units
112
Render output units
32
L1 cache
16 KB
L2 cache
0.5 MB

Theoretical performance

Peak arithmetic rates published for the board. These are ceilings that no real workload reaches, and generating text reaches a small fraction of them because it is limited by memory rather than arithmetic.

Single precision (FP32)
2 TFLOPS
Double precision (FP64)
84.9 GFLOPS
Pixel rate
21 GPixel/s
Texture rate
85 GTexel/s

The board

What it takes to physically install and power the card — the practical constraints that decide whether it fits the machine you already own.

Power draw (TDP)
100 W
Power connectors
None
Bus interface
MXM-B (3.0)
Slot width
MXM Module

Software support

Which graphics and compute interfaces the card supports. CUDA compute capability is the one that bears on inference: below 7.0 there are no tensor cores, and modern inference software falls back to slower code paths.

CUDA compute capability
3.0
DirectX
11.0
OpenGL
4.6
Vulkan
1.2
OpenCL
3.0
Shader model
5.1

Listings

Where to buy a GeForce GTX 680M

No vendor is currently listing this card. Listings come from vendors who publish them here directly — browse the vendor directory to see who is selling what.

What the numbers mean

Memory: the specification that decides everything

Memory

4 GB

Bandwidth

115 GB/s

Largest model

DeciLM 6B

At 4 GB of GDDR5 the GeForce GTX 680M is limited to the smaller end of the catalogue. About 3.6 GB is actually available to a runtime, and a model has to fit entirely inside it before generating anything at all.

At 115 GB/s across a 256-bit bus, bandwidth is this card's real constraint. Every token requires reading the entire model out of memory, so a large model will feel slow here even when it fits.

The figure is the memory clock — 900 MHz here — multiplied by the bus width. It is why core counts predict generation speed so poorly.

The practical ceiling is DeciLM 6B at 5.7B, held at Q3_K_M and running at roughly 19.6 tokens per second.

The chip and how it was built

The GeForce GTX 680M is built on the GK104 graphics processor, using NVIDIA's Kepler architecture, as part of the GeForce 600M generation.

The chip is manufactured by TSMC, on a 28 nm process, with a die measuring 294 mm², holding 3.5 billion transistors. A smaller process generally means more performance for the same power, though for language models it matters far less than the memory subsystem.

It was released in June 2012, roughly 14 years ago. Inference software support tends to follow hardware by a year or two, so a card of this age generally has mature, well-optimised code paths available to it.

Compute throughput, and why it matters less than it looks

FP64

84.9 GFLOPS

Double-precision throughput is 84.9 GFLOPS. It has no bearing on running a language model — no inference runtime uses it — but it separates datacentre parts from consumer ones, since the latter deliberately restrict it.

Clocks run from 719 MHz at base to 758 MHz boosted. Worth far less here than on a gaming benchmark: raising the clock speeds up the arithmetic, and the arithmetic is not what generation is waiting on.

Cache and processing units

The GeForce GTX 680M has 16 KB of L1 cache, backed by 0.5 MB of L2. Cache absorbs a share of the memory traffic that would otherwise hit the main bus, which is the one place on this page where a number other than bandwidth quietly affects generation speed — a large L2 lets more of the working set stay close to the cores.

There are 1,344 shading units, 112 texture mapping units, and 32 render output units. These drive graphics workloads and contribute to prompt processing, but they sit idle for much of the time a model spends generating a reply.

Power, size and installation

Power draw

100 W

The GeForce GTX 680M is rated at 100 W. Running a language model keeps a card busy in bursts rather than continuously — it draws hard while generating and idles between requests — so sustained draw over a working day is usually well below the rated figure.

The board occupies a mxm module. Worth checking against the case and power supply already in the machine, since the largest cards need considerably more of both than a typical desktop provides.

It connects over MXM-B (3.0). The interface governs how quickly a model is loaded from disk into the card, not how fast it runs once there, so a narrower link costs a few seconds at startup and nothing thereafter.

The extremes

The largest AI models that run on a GeForce GTX 680M

The biggest open-weight models that fit on this card, newest first. Each is shown at the best compression the card can hold.

  1. 01 Qwen3.5-4B 4B · Q5_K_M · Feb 2026 18.5 tok/s
  2. 02 Voxtral Mini 4.7B · Q4_K_M · Jul 2025 20.4 tok/s
  3. 03 Typhoon 2.1 Gemma 4B 4B · Q4_K_M · May 2025 23.9 tok/s
  4. 04 Qwen3-4B 4B · Q3_K_M · Apr 2025 28.0 tok/s
  5. 05 Gemma 3 QAT 4B 4B · Q4_K_M · Apr 2025 23.9 tok/s
  6. 06 Gemma 3 4B 4B · Q4_K_M · Mar 2025 23.9 tok/s
  7. 07 Phi-4-Multimodal 5.6B · Q3_K_M · Mar 2025 20.0 tok/s
  8. 08 Minitron 4B 4.2B · Q4_K_M · Nov 2024 22.8 tok/s
  9. 09 XVERSE-MoE-A4.2B 4.2B · Q4_K_M · Apr 2024 22.8 tok/s
  10. 10 DeciLM 6B 5.7B · Q3_K_M · Sep 2023 19.6 tok/s

The fastest AI models on a GeForce GTX 680M

Where this card produces tokens quickest. Smaller models dominate here, because generating each token means reading the whole model out of memory once.

  1. 01 Gemma 3 QAT 1B 1B · Q8_0 · 1.8 GB 41.5 tok/s
  2. 02 Gemma 3 1B 1B · Q8_0 · 1.8 GB 41.5 tok/s
  3. 03 LLama 3..2 Typhoon 2 1B 1B · Q8_0 · 1.8 GB 41.5 tok/s
  4. 04 OLMo-1B 1B · Q8_0 · 1.8 GB 41.5 tok/s
  5. 05 HGRN 1B (WT 103) 1B · Q8_0 · 1.8 GB 41.5 tok/s
  6. 06 Pythia-1b 1B · Q8_0 · 1.8 GB 41.5 tok/s
  7. 07 OpenELM-1.1B 1.1B · Q8_0 · 1.9 GB 38.4 tok/s
  8. 08 TinyLlama-1.1B (1T token checkpoint) 1.1B · Q8_0 · 1.9 GB 37.7 tok/s
  9. 09 TinyLlama-1.1B (3T token checkpoint) 1.1B · Q8_0 · 1.9 GB 37.7 tok/s
  10. 10 DeciCoder-1B 1.1B · Q8_0 · 1.9 GB 37.7 tok/s

Step by step

How to work out the tokens per second of a GeForce GTX 680M

You do not have to calculate anything by hand — the gputps.com calculator on this page has already worked it out for every model this card can hold. Reading off the answer takes six steps.

  1. 01

    Search for the model you want

    All 97 models the GeForce GTX 680M handles are already listed. The search box takes a name or a size such as 27b, which matches on parameter count.

  2. 02

    Decide how long your conversations run

    Set the context to your real working length. Short questions cost almost nothing; a long document can consume a large share of the card's 4 GB.

  3. 03

    Set a minimum quality if you need one

    Each model is shown at the best compression this card can hold. A minimum quality hides the ones that only fit by being squeezed further than you would accept.

  4. 04

    Read the speed and the range

    Speeds come with error bars for a reason. The best case here is 41.5 tok/s on Gemma 3 QAT 1B, and which inference software you use moves that by thirty to fifty per cent.

  5. 05

    Check the headroom before you decide

    Compare what each model needs with the 4 GB this card provides. Tight means it works today; comfortable means it still works when the conversation grows.

  6. 06

    Cross-check against other hardware

    Every model name in the table links to its own page, which runs the same calculation across every card we hold. That is where you see whether the GeForce GTX 680M is the right buy for it or merely a card that fits.

Answers

GeForce GTX 680M — common questions

01

Can a GeForce GTX 680M run a model that does not fit in its memory?

Offloading past the card's 4 GB is possible and usually a false economy: the system-memory portion is slow enough to dominate the result.

02

Would two GeForce GTX 680M cards be twice as fast?

Pairing GeForce GTX 680M cards buys headroom rather than pace: 8 GB of combined memory, at roughly the same generation speed as one.

03

What AI models can a GeForce GTX 680M run?

97 of the 679 open-weight language models we track fit on a GeForce GTX 680M and can be run locally on it. The table on this page lists every one, with the memory it needs, the quantisation it runs at and an estimated generation speed.

04

What is the largest AI model a GeForce GTX 680M can run?

The largest model in our catalogue that fits on a GeForce GTX 680M is DeciLM 6B at 5.7B parameters, compressed to Q3_K_M. It generates roughly 19.6 tokens per second and needs about 3.5 GB of the card's memory.

05

How many tokens per second does a GeForce GTX 680M produce?

It depends on the model. On a GeForce GTX 680M the fastest model we track is Gemma 3 QAT 1B at about 41.5 tokens per second, while larger models run proportionally slower because each token requires reading the whole model out of memory once. Speeds are estimates for a single conversation at a time.

06

How much memory does a GeForce GTX 680M have?

A GeForce GTX 680M has 4 GB of GDDR5 memory. Around a tenth of that is reserved by the inference runtime and the driver, leaving roughly 3.6 GB available for a model and its conversation.

07

What is the memory bandwidth of a GeForce GTX 680M?

The GeForce GTX 680M has 115 GB/s of memory bandwidth, across a 256-bit memory bus. This is the single best predictor of how fast it generates text, because producing each token means reading the entire model out of memory once.

08

What type of memory does a GeForce GTX 680M use?

It uses GDDR5 clocked at 900 MHz. HBM types are found on datacentre accelerators and carry far more bandwidth than the GDDR used on desktop cards, which is why they generate tokens considerably faster at the same capacity.

09

Who makes the GeForce GTX 680M?

The GeForce GTX 680M is a NVIDIA product, with the chip manufactured by TSMC, on a 28 nm process.

10

When was the GeForce GTX 680M released?

The GeForce GTX 680M was released in June 2012.

11

How much power does a GeForce GTX 680M use?

The GeForce GTX 680M has a rated board power of 100 W. Generating text draws hard in bursts and idles between requests, so average consumption over a working session is normally well below the rated figure.

12

How much cache does a GeForce GTX 680M have?

The GeForce GTX 680M has 16 KB of L1 cache, and 0.5 MB of L2 cache. Cache absorbs part of the memory traffic that would otherwise reach the main bus, so a larger L2 gives a modest lift to generation speed beyond what bandwidth alone predicts.

13

Does the GeForce GTX 680M support CUDA?

Yes. The GeForce GTX 680M reports CUDA compute capability 3.0, which predates tensor cores. Capability 7.0 and above has tensor cores, which modern inference software uses; below that it falls back to slower code paths for quantised models.

14

What bus interface does the GeForce GTX 680M use?

It uses MXM-B (3.0). This governs how fast a model is loaded onto the card rather than how fast it runs once loaded, so it costs a few seconds at startup and nothing during generation.

15

Is the GeForce GTX 680M good for running local AI models?

Its memory limits it to smaller models though its bandwidth means generation will feel slow on larger models. In total it runs 97 of the models we track. Whether that is enough depends entirely on which model you want — the table above answers that directly.

The other direction

Looking at it from the other side?

This page starts from the hardware. If you already know which model you want and need to know what it takes to run it, start from the model instead.

All GPUs