Qwen

Qwen2.5 7B Instruct

Tool callingStructured output
Qwen2.5 7B Instruct model card cover
Context
33K
Released
2y ago
Parameters
7B
Input / 1M tokens
$0.10
Output / 1M tokens
$0.20

Sources:OpenRouter

Benchmarks

DeepSWE is the primary coding benchmark where an exact model evaluation exists. Other sourced evaluations appear below; missing scores are not zero.

Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.

Other published benchmarks

EvaluationReported score and source
GPQA diamond
35.5%
Epoch AI · qwen2.5-7b-instruct
Epoch AI evaluation
OTIS Mock AIME 2024-2025
2.5%
Epoch AI · qwen2.5-7b-instruct
Epoch AI evaluation

Benchmark compilation: Epoch AI, Capabilities & benchmarking (CC BY attribution). Scores are converted to percentages and model names matched to the Vertical Tabs catalog. Individual evaluations are credited to their original sources. External results may carry separate licences and evaluation configurations; do not interpret scores from different harnesses as directly comparable.

Specifications

Developer
Qwen
Model family
Qwen
Type
Language model
Predecessor
—
Successor
—
Released
September 19, 2024
Model ID
qwen/qwen-2.5-7b-instruct
Openness
Open weights
License
—
Parameters
7B
Input modalities
text
Output modalities
text
Modality
text->text
Native context
32K
Tokenizer
Qwen
Max output tokens
29K
Knowledge cutoff
2024-06-30
Source freshness
Endpoints synced Oct 15, 2026, 8:17 AM UTC

Open-weight builds

Ollama variants

Every published Ollama tag matched to this model, including parameter editions, instruction/base variants, and quantized builds.

117 builds
VariantParametersTypeQuantizationSizeContextInput
qwen2.5:0.5b-base-q2_K0.5BbaseQ2_K339 MB32Ktext
qwen2.5:0.5b-base-q3_K_L0.5BbaseQ3_K_L369 MB32Ktext
qwen2.5:0.5b-base-q3_K_M0.5BbaseQ3_K_M355 MB32Ktext
qwen2.5:0.5b-base-q3_K_S0.5BbaseQ3_K_S338 MB32Ktext
qwen2.5:0.5b-base-q4_00.5BbaseQ4_0352 MB32Ktext
qwen2.5:0.5b-base-q4_10.5BbaseQ4_1375 MB32Ktext
qwen2.5:0.5b-base0.5BbaseQ4_K_M398 MB32Ktext
qwen2.5:0.5b-base-q4_K_S0.5BbaseQ4_K_S385 MB32Ktext
qwen2.5:0.5b-base-q5_00.5BbaseQ5_0397 MB32Ktext
qwen2.5:0.5b-base-q5_10.5BbaseQ5_1419 MB32Ktext
qwen2.5:0.5b-base-q5_K_S0.5BbaseQ5_K_S413 MB32Ktext
qwen2.5:0.5b-base-q8_00.5BbaseQ8_0531 MB32Ktext
qwen2.5:0.5b-instruct-fp160.5BinstructFP16994 MB32Ktext
qwen2.5:0.5b-instruct-q2_K0.5BinstructQ2_K339 MB32Ktext
qwen2.5:0.5b-instruct-q3_K_L0.5BinstructQ3_K_L369 MB32Ktext
qwen2.5:0.5b-instruct-q3_K_M0.5BinstructQ3_K_M355 MB32Ktext
qwen2.5:0.5b-instruct-q3_K_S0.5BinstructQ3_K_S338 MB32Ktext
qwen2.5:0.5b-instruct-q4_00.5BinstructQ4_0352 MB32Ktext
qwen2.5:0.5b-instruct-q4_10.5BinstructQ4_1375 MB32Ktext
qwen2.5:0.5b0.5BinstructQ4_K_M398 MB32Ktext
qwen2.5:0.5b-instruct-q4_K_S0.5BinstructQ4_K_S385 MB32Ktext
qwen2.5:0.5b-instruct-q5_00.5BinstructQ5_0397 MB32Ktext
qwen2.5:0.5b-instruct-q5_10.5BinstructQ5_1419 MB32Ktext
qwen2.5:0.5b-instruct-q5_K_M0.5BinstructQ5_K_M420 MB32Ktext
qwen2.5:0.5b-instruct-q5_K_S0.5BinstructQ5_K_S413 MB32Ktext
qwen2.5:0.5b-instruct-q6_K0.5BinstructQ6_K506 MB32Ktext
qwen2.5:0.5b-instruct-q8_00.5BinstructQ8_0531 MB32Ktext
qwen2.5:1.5b-instruct-fp161.5BinstructFP163.1 GB32Ktext
qwen2.5:1.5b-instruct-q2_K1.5BinstructQ2_K676 MB32Ktext
qwen2.5:1.5b-instruct-q3_K_L1.5BinstructQ3_K_L880 MB32Ktext
qwen2.5:1.5b-instruct-q3_K_M1.5BinstructQ3_K_M824 MB32Ktext
qwen2.5:1.5b-instruct-q3_K_S1.5BinstructQ3_K_S761 MB32Ktext
qwen2.5:1.5b-instruct-q4_01.5BinstructQ4_0935 MB32Ktext
qwen2.5:1.5b-instruct-q4_11.5BinstructQ4_11 GB32Ktext
qwen2.5:1.5b1.5BinstructQ4_K_M986 MB32Ktext
qwen2.5:1.5b-instruct-q4_K_S1.5BinstructQ4_K_S940 MB32Ktext
qwen2.5:1.5b-instruct-q5_01.5BinstructQ5_01.1 GB32Ktext
qwen2.5:1.5b-instruct-q5_11.5BinstructQ5_11.2 GB32Ktext
qwen2.5:1.5b-instruct-q5_K_M1.5BinstructQ5_K_M1.1 GB32Ktext
qwen2.5:1.5b-instruct-q5_K_S1.5BinstructQ5_K_S1.1 GB32Ktext
qwen2.5:1.5b-instruct-q6_K1.5BinstructQ6_K1.3 GB32Ktext
qwen2.5:1.5b-instruct-q8_01.5BinstructQ8_01.6 GB32Ktext
qwen2.5:3b-instruct-fp163BinstructFP166.2 GB32Ktext
qwen2.5:3b-instruct-q2_K3BinstructQ2_K1.3 GB32Ktext
qwen2.5:3b-instruct-q3_K_L3BinstructQ3_K_L1.7 GB32Ktext
qwen2.5:3b-instruct-q3_K_M3BinstructQ3_K_M1.6 GB32Ktext
qwen2.5:3b-instruct-q3_K_S3BinstructQ3_K_S1.5 GB32Ktext
qwen2.5:3b-instruct-q4_03BinstructQ4_01.8 GB32Ktext
qwen2.5:3b-instruct-q4_13BinstructQ4_12 GB32Ktext
qwen2.5:3b3BinstructQ4_K_M1.9 GB32Ktext
qwen2.5:3b-instruct-q4_K_S3BinstructQ4_K_S1.8 GB32Ktext
qwen2.5:3b-instruct-q5_03BinstructQ5_02.2 GB32Ktext
qwen2.5:3b-instruct-q5_13BinstructQ5_12.3 GB32Ktext
qwen2.5:3b-instruct-q5_K_M3BinstructQ5_K_M2.2 GB32Ktext
qwen2.5:3b-instruct-q5_K_S3BinstructQ5_K_S2.2 GB32Ktext
qwen2.5:3b-instruct-q6_K3BinstructQ6_K2.5 GB32Ktext
qwen2.5:3b-instruct-q8_03BinstructQ8_03.3 GB32Ktext
qwen2.5:7b-instruct-fp167BinstructFP1615 GB32Ktext
qwen2.5:7b-instruct-q2_K7BinstructQ2_K3 GB32Ktext
qwen2.5:7b-instruct-q3_K_L7BinstructQ3_K_L4.1 GB32Ktext
qwen2.5:7b-instruct-q3_K_M7BinstructQ3_K_M3.8 GB32Ktext
qwen2.5:7b-instruct-q3_K_S7BinstructQ3_K_S3.5 GB32Ktext
qwen2.5:7b-instruct-q4_07BinstructQ4_04.4 GB32Ktext
qwen2.5:7b-instruct-q4_17BinstructQ4_14.9 GB32Ktext
qwen2.5:latest7BinstructQ4_K_M4.7 GB32Ktext
qwen2.5:7b-instruct-q4_K_S7BinstructQ4_K_S4.5 GB32Ktext
qwen2.5:7b-instruct-q5_07BinstructQ5_05.3 GB32Ktext
qwen2.5:7b-instruct-q5_17BinstructQ5_15.8 GB32Ktext
qwen2.5:7b-instruct-q5_K_M7BinstructQ5_K_M5.4 GB32Ktext
qwen2.5:7b-instruct-q5_K_S7BinstructQ5_K_S5.3 GB32Ktext
qwen2.5:7b-instruct-q6_K7BinstructQ6_K6.3 GB32Ktext
qwen2.5:7b-instruct-q8_07BinstructQ8_08.1 GB32Ktext
qwen2.5:14b-instruct-fp1614BinstructFP1630 GB32Ktext
qwen2.5:14b-instruct-q2_K14BinstructQ2_K5.8 GB32Ktext
qwen2.5:14b-instruct-q3_K_L14BinstructQ3_K_L7.9 GB32Ktext
qwen2.5:14b-instruct-q3_K_M14BinstructQ3_K_M7.3 GB32Ktext
qwen2.5:14b-instruct-q3_K_S14BinstructQ3_K_S6.7 GB32Ktext
qwen2.5:14b-instruct-q4_014BinstructQ4_08.5 GB32Ktext
qwen2.5:14b-instruct-q4_114BinstructQ4_19.4 GB32Ktext
qwen2.5:14b14BinstructQ4_K_M9 GB32Ktext
qwen2.5:14b-instruct-q4_K_S14BinstructQ4_K_S8.6 GB32Ktext
qwen2.5:14b-instruct-q5_014BinstructQ5_010 GB32Ktext
qwen2.5:14b-instruct-q5_114BinstructQ5_111 GB32Ktext
qwen2.5:14b-instruct-q5_K_M14BinstructQ5_K_M11 GB32Ktext
qwen2.5:14b-instruct-q5_K_S14BinstructQ5_K_S10 GB32Ktext
qwen2.5:14b-instruct-q6_K14BinstructQ6_K12 GB32Ktext
qwen2.5:14b-instruct-q8_014BinstructQ8_016 GB32Ktext
qwen2.5:32b-instruct-fp1632BinstructFP1666 GB32Ktext
qwen2.5:32b-instruct-q2_K32BinstructQ2_K12 GB32Ktext
qwen2.5:32b-instruct-q3_K_L32BinstructQ3_K_L17 GB32Ktext
qwen2.5:32b-instruct-q3_K_M32BinstructQ3_K_M16 GB32Ktext
qwen2.5:32b-instruct-q3_K_S32BinstructQ3_K_S14 GB32Ktext
qwen2.5:32b-instruct-q4_032BinstructQ4_019 GB32Ktext
qwen2.5:32b-instruct-q4_132BinstructQ4_121 GB32Ktext
qwen2.5:32b32BinstructQ4_K_M20 GB32Ktext
qwen2.5:32b-instruct-q4_K_S32BinstructQ4_K_S19 GB32Ktext
qwen2.5:32b-instruct-q5_032BinstructQ5_023 GB32Ktext
qwen2.5:32b-instruct-q5_132BinstructQ5_125 GB32Ktext
qwen2.5:32b-instruct-q5_K_M32BinstructQ5_K_M23 GB32Ktext
qwen2.5:32b-instruct-q5_K_S32BinstructQ5_K_S23 GB32Ktext
qwen2.5:32b-instruct-q6_K32BinstructQ6_K27 GB32Ktext
qwen2.5:32b-instruct-q8_032BinstructQ8_035 GB32Ktext
qwen2.5:72b-instruct-fp1672BinstructFP16145 GB32Ktext
qwen2.5:72b-instruct-q2_K72BinstructQ2_K30 GB32Ktext
qwen2.5:72b-instruct-q3_K_L72BinstructQ3_K_L40 GB32Ktext
qwen2.5:72b-instruct-q3_K_M72BinstructQ3_K_M38 GB32Ktext
qwen2.5:72b-instruct-q3_K_S72BinstructQ3_K_S34 GB32Ktext
qwen2.5:72b-instruct-q4_072BinstructQ4_041 GB32Ktext
qwen2.5:72b-instruct-q4_172BinstructQ4_146 GB32Ktext
qwen2.5:72b72BinstructQ4_K_M47 GB32Ktext
qwen2.5:72b-instruct-q4_K_S72BinstructQ4_K_S44 GB32Ktext
qwen2.5:72b-instruct-q5_072BinstructQ5_050 GB32Ktext
qwen2.5:72b-instruct-q5_172BinstructQ5_155 GB32Ktext
qwen2.5:72b-instruct-q5_K_M72BinstructQ5_K_M54 GB32Ktext
qwen2.5:72b-instruct-q5_K_S72BinstructQ5_K_S51 GB32Ktext
qwen2.5:72b-instruct-q6_K72BinstructQ6_K64 GB32Ktext
qwen2.5:72b-instruct-q8_072BinstructQ8_077 GB32Ktext

Infrastructure

Provider availability

All providers
ProviderEndpointContextInput /M tokensOutput /M tokensCached input /M tokensCache write /M tokensUptime 1dData handling
Phalaphala33K$0.10$0.20——100.00%
ZDR: Yes

Training: Unknown

Prompt logging: Unknown

Retention: Unknown

Checked 2026-10-15

Provider data:OpenRouter

Deployment graph

Can I run this?

Vertical Tabs only shows compatibility when it has structured evidence tying this exact model to hardware, runtime, quantization, memory use, and observed performance.

Browse hardware