Viewing enriched Twitter post
Fish Audio just benchmarked SGLang, vLLM, and MAX 👀 TLDR: 16% faster throughput than vLLM on L40, p99 TTFT of 13.1ms vs 23.6ms, containers under 700MB. The only stack in the comparison built without CUDA, running across NVIDIA, AMD, Apple Silicon, and CPU from one codebase. https://t.co/JAE4e69agh
{
"media": [
{
"url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2044140427327221791/media_0.jpg",
"media_url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2044140427327221791/media_0.jpg",
"type": "photo",
"filename": "media_0.jpg"
}
],
"processed_at": "2026-04-14T19:54:05.676413",
"pipeline_version": "2.0"
}{
"type": "tweet",
"id": "2044140427327221791",
"url": "https://x.com/Modular/status/2044140427327221791",
"twitterUrl": "https://twitter.com/Modular/status/2044140427327221791",
"text": "Fish Audio just benchmarked SGLang, vLLM, and MAX 👀 TLDR: 16% faster throughput than vLLM on L40, p99 TTFT of 13.1ms vs 23.6ms, containers under 700MB.\n\nThe only stack in the comparison built without CUDA, running across NVIDIA, AMD, Apple Silicon, and CPU from one codebase.\n\nhttps://t.co/JAE4e69agh",
"source": "Twitter for iPhone",
"retweetCount": 1,
"replyCount": 0,
"likeCount": 4,
"quoteCount": 0,
"viewCount": 120,
"createdAt": "Tue Apr 14 19:47:10 +0000 2026",
"lang": "en",
"bookmarkCount": 1,
"isReply": false,
"inReplyToId": null,
"conversationId": "2044140427327221791",
"displayTextRange": [
0,
275
],
"inReplyToUserId": null,
"inReplyToUsername": null,
"author": {
"type": "user",
"userName": "Modular",
"url": "https://x.com/Modular",
"twitterUrl": "https://twitter.com/Modular",
"id": "1483918307484848132",
"name": "Modular",
"isVerified": false,
"isBlueVerified": true,
"verifiedType": "Business",
"profilePicture": "https://pbs.twimg.com/profile_images/1786875652656025600/mp6VADd5_normal.png",
"coverPicture": "https://pbs.twimg.com/profile_banners/1483918307484848132/1714859329",
"description": "Building AI’s unified compute layer. We are hiring → https://t.co/cPTAes0HMt 🚀",
"location": "",
"followers": 22313,
"following": 2,
"status": "",
"canDm": false,
"canMediaTag": true,
"createdAt": "Wed Jan 19 21:46:38 +0000 2022",
"entities": {
"description": {
"urls": [
{
"display_url": "modular.com/careers",
"expanded_url": "http://modular.com/careers",
"indices": [
53,
76
],
"url": "https://t.co/cPTAes0HMt"
}
]
},
"url": {
"urls": [
{
"display_url": "modular.com",
"expanded_url": "https://www.modular.com",
"indices": [
0,
23
],
"url": "https://t.co/dFAH0NVA0N"
}
]
}
},
"fastFollowersCount": 0,
"favouritesCount": 697,
"hasCustomTimelines": false,
"isTranslator": false,
"mediaCount": 350,
"statusesCount": 1125,
"withheldInCountries": [],
"affiliatesHighlightedLabel": {},
"possiblySensitive": false,
"pinnedTweetIds": [
"2039738561436713330"
],
"profile_bio": {},
"isAutomated": false,
"automatedBy": null
},
"extendedEntities": {},
"card": {
"binding_values": [
{
"key": "photo_image_full_size_large",
"value": {
"image_value": {
"height": 419,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=800x419",
"width": 800
},
"type": "IMAGE"
}
},
{
"key": "thumbnail_image",
"value": {
"image_value": {
"height": 150,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=280x150",
"width": 269
},
"type": "IMAGE"
}
},
{
"key": "description",
"value": {
"string_value": "As AI models move from research to production, the inference engine you choose determines your latency, throughput, and infrastructure cost. The open-source eco",
"type": "STRING"
}
},
{
"key": "domain",
"value": {
"string_value": "fish.audio",
"type": "STRING"
}
},
{
"key": "thumbnail_image_large",
"value": {
"image_value": {
"height": 320,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=800x320_1",
"width": 573
},
"type": "IMAGE"
}
},
{
"key": "summary_photo_image_small",
"value": {
"image_value": {
"height": 202,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=386x202",
"width": 386
},
"type": "IMAGE"
}
},
{
"key": "thumbnail_image_original",
"value": {
"image_value": {
"height": 768,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=orig",
"width": 1376
},
"type": "IMAGE"
}
},
{
"key": "photo_image_full_size_small",
"value": {
"image_value": {
"height": 202,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=386x202",
"width": 386
},
"type": "IMAGE"
}
},
{
"key": "summary_photo_image_large",
"value": {
"image_value": {
"height": 419,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=800x419",
"width": 800
},
"type": "IMAGE"
}
},
{
"key": "thumbnail_image_small",
"value": {
"image_value": {
"height": 80,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=144x144",
"width": 144
},
"type": "IMAGE"
}
},
{
"key": "thumbnail_image_x_large",
"value": {
"image_value": {
"height": 768,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=png&name=2048x2048_2_exp",
"width": 1376
},
"type": "IMAGE"
}
},
{
"key": "photo_image_full_size_original",
"value": {
"image_value": {
"height": 768,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=orig",
"width": 1376
},
"type": "IMAGE"
}
},
{
"key": "photo_image_full_size_alt_text",
"value": {
"string_value": "Open-source LLM inference engines compared: SGLang, vLLM, MAX, and BentoML 2026",
"type": "STRING"
}
},
{
"key": "vanity_url",
"value": {
"scribe_key": "vanity_url",
"string_value": "fish.audio",
"type": "STRING"
}
},
{
"key": "photo_image_full_size",
"value": {
"image_value": {
"height": 314,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=600x314",
"width": 600
},
"type": "IMAGE"
}
},
{
"key": "summary_photo_image_alt_text",
"value": {
"string_value": "Open-source LLM inference engines compared: SGLang, vLLM, MAX, and BentoML 2026",
"type": "STRING"
}
},
{
"key": "thumbnail_image_color",
"value": {
"image_color_value": {
"palette": [
{
"percentage": 91.84,
"rgb": {
"blue": 233,
"green": 245,
"red": 249
}
},
{
"percentage": 3.39,
"rgb": {
"blue": 107,
"green": 158,
"red": 210
}
},
{
"percentage": 2.57,
"rgb": {
"blue": 184,
"green": 175,
"red": 205
}
},
{
"percentage": 1.75,
"rgb": {
"blue": 126,
"green": 136,
"red": 201
}
},
{
"percentage": 0.35,
"rgb": {
"blue": 190,
"green": 235,
"red": 249
}
}
]
},
"type": "IMAGE_COLOR"
}
},
{
"key": "title",
"value": {
"string_value": "Open-source LLM inference engines compared: SGLang, vLLM, MAX, and BentoML 2026",
"type": "STRING"
}
},
{
"key": "summary_photo_image_color",
"value": {
"image_color_value": {
"palette": [
{
"percentage": 91.84,
"rgb": {
"blue": 233,
"green": 245,
"red": 249
}
},
{
"percentage": 3.39,
"rgb": {
"blue": 107,
"green": 158,
"red": 210
}
},
{
"percentage": 2.57,
"rgb": {
"blue": 184,
"green": 175,
"red": 205
}
},
{
"percentage": 1.75,
"rgb": {
"blue": 126,
"green": 136,
"red": 201
}
},
{
"percentage": 0.35,
"rgb": {
"blue": 190,
"green": 235,
"red": 249
}
}
]
},
"type": "IMAGE_COLOR"
}
},
{
"key": "summary_photo_image_x_large",
"value": {
"image_value": {
"height": 768,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=png&name=2048x2048_2_exp",
"width": 1376
},
"type": "IMAGE"
}
},
{
"key": "summary_photo_image",
"value": {
"image_value": {
"height": 314,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=600x314",
"width": 600
},
"type": "IMAGE"
}
},
{
"key": "photo_image_full_size_color",
"value": {
"image_color_value": {
"palette": [
{
"percentage": 91.84,
"rgb": {
"blue": 233,
"green": 245,
"red": 249
}
},
{
"percentage": 3.39,
"rgb": {
"blue": 107,
"green": 158,
"red": 210
}
},
{
"percentage": 2.57,
"rgb": {
"blue": 184,
"green": 175,
"red": 205
}
},
{
"percentage": 1.75,
"rgb": {
"blue": 126,
"green": 136,
"red": 201
}
},
{
"percentage": 0.35,
"rgb": {
"blue": 190,
"green": 235,
"red": 249
}
}
]
},
"type": "IMAGE_COLOR"
}
},
{
"key": "photo_image_full_size_x_large",
"value": {
"image_value": {
"height": 768,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=png&name=2048x2048_2_exp",
"width": 1376
},
"type": "IMAGE"
}
},
{
"key": "card_url",
"value": {
"scribe_key": "card_url",
"string_value": "https://t.co/JAE4e69agh",
"type": "STRING"
}
},
{
"key": "summary_photo_image_original",
"value": {
"image_value": {
"height": 768,
"url": "https://pbs.twimg.com/card_img/2044140431995490304/Iu5sCml0?format=jpg&name=orig",
"width": 1376
},
"type": "IMAGE"
}
}
],
"card_platform": {
"platform": {
"audience": {
"name": "production"
},
"device": {
"name": "Swift",
"version": "12"
}
}
},
"name": "summary_large_image",
"url": "https://t.co/JAE4e69agh",
"user_refs_results": []
},
"place": {},
"entities": {
"hashtags": [],
"symbols": [],
"urls": [
{
"display_url": "fish.audio/blog/open-sour…",
"expanded_url": "https://fish.audio/blog/open-source-llm-inference-engines-2026/",
"indices": [
277,
300
],
"url": "https://t.co/JAE4e69agh"
}
],
"user_mentions": []
},
"quoted_tweet": null,
"retweeted_tweet": null,
"article": null
}