# Rapidata General image — coherence > Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/rapidata%3Aimage%3Arlb_1PetHDuThgAvpj. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/rapidata%3Aimage%3Arlb_1PetHDuThgAvpj. Human pairwise evaluation of text-to-image outputs, measuring coherence. Scores are fitted using the Bradley–Terry preference model and expressed on the publisher’s Elo scale with 95% confidence intervals. Canonical URL: https://benchize.com/benchmarks/rapidata%3Aimage%3Arlb_1PetHDuThgAvpj Last updated: Wed, 21 Oct 2026 18:52:20 GMT Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations. Human pairwise evaluation of text-to-image outputs, measuring coherence. Scores are fitted using the Bradley–Terry preference model and expressed on the publisher’s Elo scale with 95% confidence intervals. Multimodal · Models tested: 12 · Score publishers: [rapidata](https://www.benchmark.ai/image) ## Model results | # | Model | Score | ± | | | --- | --- | --- | --- | --- | | 1 | [GPT Image 2.5 Sunburst](https://benchize.com/models/gpt-image-2-5-sunburst.md) | 1,246.9elo | CI 1,236.93–1,256.88 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1246.9044379256525 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1Ub5xBI0tHZKcM sourceModelVersion GPT Image 2.5 (Sunburst) weighted false ### Reported statistics Uncertainty Confidence interval: 1,236.93–1,256.88 confidenceIntervalLowerOffset 9.973883513718192 confidenceIntervalUpperOffset 9.973883513718192 totalMatches 27555 wins 18089.5 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label GPT Image 2.5 (Sunburst) Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/0` | | | | | | 2 | [GPT Image 2.5 Flare](https://benchize.com/models/gpt-image-2-5-flare.md) | 1,218.73elo | CI 1,208.83–1,228.64 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1218.7346419101113 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1UamSKYVEFc6ND sourceModelVersion GPT Image 2.5 (Flare) weighted false ### Reported statistics Uncertainty Confidence interval: 1,208.83–1,228.64 confidenceIntervalLowerOffset 9.906443563219682 confidenceIntervalUpperOffset 9.906443563219682 totalMatches 27354 wins 17514.5 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label GPT Image 2.5 (Flare) Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/1` | | | | | | 3 | [Nano Banana 2.1](https://benchize.com/models/gemini-nano-banana-2-1.md) | 1,138.41elo | CI 1,110.62–1,166.19 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1138.405231376729 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1XDkhTdAbFixop sourceModelVersion nano-banana-2.1 weighted false ### Reported statistics Uncertainty Confidence interval: 1,110.62–1,166.19 confidenceIntervalLowerOffset 27.78229889226759 confidenceIntervalUpperOffset 27.782298892267818 totalMatches 3198 wins 1872 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label nano-banana-2.1 Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/4` | | | | | | 4 | [Recraft V4.1 Utility Pro](https://benchize.com/models/recraft-v4-1-utility-pro.md) | 1,003.9elo | CI 992.45–1,015.34 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1003.8987615349836 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1PetVSiud8YtIo sourceModelVersion Recraft V4.1 Utility Pro weighted false ### Reported statistics Uncertainty Confidence interval: 992.45–1,015.34 confidenceIntervalLowerOffset 11.444517220207103 confidenceIntervalUpperOffset 11.44451722020699 totalMatches 18734 wins 9428.5 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label Recraft V4.1 Utility Pro Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/24` | | | | | | 5 | [Qwen Image](https://benchize.com/models/qwen-image.md) | 1,002.88elo | CI 991.82–1,013.94 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1002.8778371262061 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1Petln7VHOEQLz sourceModelVersion Qwen-Image weighted false ### Reported statistics Uncertainty Confidence interval: 991.82–1,013.94 confidenceIntervalLowerOffset 11.058973164223517 confidenceIntervalUpperOffset 11.058973164223517 totalMatches 20090 wins 10172 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label Qwen-Image Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/25` | | | | | | 6 | [Flux 2 Pro](https://benchize.com/models/flux-2-pro.md) | 991.74elo | CI 980.29–1,003.19 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 991.743029243206 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1PetS5NORduFAr sourceModelVersion FLUX.2 pro weighted false ### Reported statistics Uncertainty Confidence interval: 980.29–1,003.19 confidenceIntervalLowerOffset 11.449460960878582 confidenceIntervalUpperOffset 11.449460960878469 totalMatches 18731 wins 9283.5 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label FLUX.2 pro Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/26` | | | | | | 7 | [FLUX.2 Max](https://benchize.com/models/flux-2-max.md) | 973.49elo | CI 962.06–984.93 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 973.4942934245272 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1PetpMtaBx0dSh sourceModelVersion FLUX.2 max weighted false ### Reported statistics Uncertainty Confidence interval: 962.06–984.93 confidenceIntervalLowerOffset 11.434756052754665 confidenceIntervalUpperOffset 11.434756052754665 totalMatches 18737 wins 9071.5 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label FLUX.2 max Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/29` | | | | | | 8 | [MAI-Image-2.5](https://benchize.com/models/mai-image-2-5.md) | 972.88elo | CI 961.42–984.35 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 972.8847262133877 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1PetshHzWmebbn sourceModelVersion MAI-Image 2.5 weighted false ### Reported statistics Uncertainty Confidence interval: 961.42–984.35 confidenceIntervalLowerOffset 11.46628272434873 confidenceIntervalUpperOffset 11.466282724348844 totalMatches 18734 wins 9063 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label MAI-Image 2.5 Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/30` | | | | | | 9 | [Seedream 4.5](https://benchize.com/models/seedream-4-5.md) | 950.03elo | CI 938.56–961.5 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 950.0322631008618 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1PeuOYZBaIPPIS sourceModelVersion Seedream 4.5 weighted false ### Reported statistics Uncertainty Confidence interval: 938.56–961.5 confidenceIntervalLowerOffset 11.469394374923695 confidenceIntervalUpperOffset 11.469394374923581 totalMatches 18738 wins 8797 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label Seedream 4.5 Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/35` | | | | | | 10 | [GPT Image 1.5](https://benchize.com/models/gpt-image-1-5-high.md)High | 896.26elo | CI 884.71–907.81 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 896.2632820153772 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1PeuL1z67NKdE6 Reasoning effort high sourceModelVersion GPT Image 1.5 (high) thinkingEnabled true weighted false ### Reported statistics Uncertainty Confidence interval: 884.71–907.81 confidenceIntervalLowerOffset 11.548433899188353 confidenceIntervalUpperOffset 11.548433899188353 totalMatches 18735 wins 8168 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label GPT Image 1.5 (high) Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/41` | | | | | | 11 | [Grok Imagine Image Quality](https://benchize.com/models/grok-imagine-image-quality.md) | 881.64elo | CI 870.06–893.22 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 881.6396583595122 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1PeuSsHW1ENeLx sourceModelVersion Grok Imagine Image (quality) weighted false ### Reported statistics Uncertainty Confidence interval: 870.06–893.22 confidenceIntervalLowerOffset 11.584080776364885 confidenceIntervalUpperOffset 11.584080776364885 totalMatches 18744 wins 8003 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label Grok Imagine Image (quality) Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/42` | | | | | | 12 | [FLUX.2 Flex](https://benchize.com/models/flux-2-flex.md) | 870.06elo | CI 858.47–881.65 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 870.0636991504686 \| elo \| rlb_1PetHDuThgAvpj \| Higher is better \| ### Configuration benchmarkId bmk_1Pet6iHzbf6RQu criterion rlb_1PetHDuThgAvpj participantId par_1Pety8kReYvYid sourceModelVersion FLUX.2 flex weighted false ### Reported statistics Uncertainty Confidence interval: 858.47–881.65 confidenceIntervalLowerOffset 11.589532614423547 confidenceIntervalUpperOffset 11.589532614423547 totalMatches 18736 wins 7865.5 ### Source [Original source](https://api.rapidata.ai/benchmark/bmk_1Pet6iHzbf6RQu/standings/query?leaderboard_id%5Beq%5D=rlb_1PetHDuThgAvpj) Source rapidata Original model label FLUX.2 flex Source version v1 Observed 2026-10-21T16:37:27.701Z Evidence SHA-256 `4e459bb2e3b046f141fd621f02e8d8eab1bd23d4d9ec60b1ce76df04f7e117e6` Source locator `/items/44` | | | | | --- Cite as: Rapidata General image — coherence — Benchize, https://benchize.com/benchmarks/rapidata%3Aimage%3Arlb_1PetHDuThgAvpj (retrieved 2026-10-21)