Ranking tekstowy

Ogólny ranking modeli do generowania tekstu, oparty na głosach społeczności i systemie ocen Elo.

Przewiń poziomo, aby zobaczyć wszystkie kolumny

Tekst

#	Model	Elo	Głosy	Benchmark	$/1M wej. / wyj.	Wartość
1	GPT-5.2	1304	140	-	$1.75 / $14.00	166
2	GLM 5	1263	165	57.0	$0.95 / $2.55	722
3	Claude Sonnet 4.6	1252	146	-	$3.00 / $15.00	139
4	Functionary Swahili Mini	1230	51	62.2	$0.10 / $0.90	2,460
5	Claude Haiku 4.5	1204	152	-	$1.00 / $5.00	401
6	Claude Sonnet 4.5	1186	128	-	$3.00 / $15.00	132
7	Functionary Swahili Large	1166	133	67.7	$0.20 / $1.50	1,372
8	GPT-oss-120B	1125	134	59.1	$0.04 / $0.17	10,870
9	GPT-5 Nano	1109	150	68.4	$0.05 / $0.40	4,929
10	Kimi K2.6	1225	5	-	$0.65 / $2.72	728
11	Step 3.7 Flash	1223	2	-	$0.20 / $1.15	1,812
12	Qwen3.7 Max	1221	6	-	$1.48 / $4.42	414
13	MiniMax M3	1217	12	-	$0.30 / $1.20	1,623
14	GLM 5.2	1210	3	-	$0.68 / $2.13	863
15	Gemini 3.5 Flash	1194	4	-	$1.50 / $9.00	227
16	DeepSeek V4 Pro	1193	2	-	$0.43 / $0.87	1,828
17	GPT-5.6 Luna	1192	1	-	$0.50 / $3.00	681
18	DeepSeek V4 Flash	1188	1	-	$0.14 / $0.28	5,657
19	Grok 4.5	1184	1	-	$2.00 / $6.00	296
20	Claude Sonnet 5	1184	4	-	$2.00 / $10.00	197