To solve the problem that the Chicken swarm optimization (CSO) has low solution accuracy and tends to fall into the local optimum on later stages of iteration, an adaptive mutation learning Chicken swarm optimization (AMLCSO) is proposed in this paper. Firstly, to solve the problem of uneven initial distribution and improve the algorithm’s stability, a good-point set is introduced. Secondly, according to the difference between the current individual position and the optimal individual position, the nonlinear adaptive adjustment of weight is realized and the position update step is dynamically adjusted. This strategy improves the algorithm’s convergence. Thirdly, the learning update strategies of Gaussian mutation and normal distribution are introduced to improve the probability of selection and solving accuracy and avoid falling into the local optimum. Finally, the AMLCSO is compared with other standard algorithms and improved Chicken swarm optimization algorithms on twenty benchmark test functions. The experimental results show the AMLCSO has faster convergence and higher solution accuracy.